YES bo ete Oe” Se Roe tad, ie entattetieatae pee es nares er eenten eo wpe rer n eel. ated * ¢ Senta recyctee aca a eth einen entr oo ees. merical naw pare ens = i en ee hear por ‘ : - 5 ee ; cae eter Te Waeteareng eee ~ ee Ren by see ate Teen mY Oe aes I a ar eres OS eae Cee = eeses Swit LE TiN OF THE AMERICAN MUSEUM OF NATURAL Tl LOK ¥. Volume VII, 1895. ST NEW YORK: PUBLISHED BY ORDER OF THE TRUSTEES, 1895. 5 i} /. 4 , = — y; it a PA: ke i 5 xi Ris in Ls i a ‘va pe u f .* 1 ina 7 ” bal : ele re \ - ba a ? pe 2 : / ta my he 7 . re’ y 7 . A ¢ j ‘% | a A. : k | et: ay. ahs Aon = jo a J rd , : j ; _ i) bt | | COMMITTEE OF PUBLICATION, | 7” i fe \ SN! Dae } t. Py | - R. P. WHITFIELD. J. A. ALLEN, ; a ; ; HENRY FAarRriELD Ossorn, WM, BrurEeNM: F. W. Putnam. 4 Me EDITOR OF BULLETIN. J, Ac: Alien ar a 4 « A. CONTENTS OF VOLUME VII. PAGE an eas Sie snag, bo ws, ofn gia AA nd Sos si SOR ILS a Ae ee ele Goumnittee ot Publicationy.s.. 2020. tes eS -coes Se roa cee pet ear ot ets 1 Ls oo Nees ea eam ao i ae iii. IPS EMPHMBULLULS LATIONS ost oy sieges rere anarchy e Me, ac d.eiats ears eos d-atajayn See's she's a oat Vs Dates of Publication of Author’s Separates......... Dua ON AE OEE Vii. Arr. I.—Fossil Mammals of the Puerco Beds. Collection of 1892. By HENRY FAIRFIELD OSBORN and CHARLES EARLE. ..... I II.—Fossil Mammals of the Uinta Basin. Expedition of 1894. By RNY sATRETETD SOSBORNE pie, «iets. c.sei jee (ole) cise Aa COIS Oe COS cers Ice Bee ay 70S Wpnonanmiaarts. Crown, VIEW Of MOlATS: ...: Side view and base view of skull... ...... on dane eae 370 Aceratherium occidentale. Crown view of fourth upper premolars ... ... 372 Tllustrations. Vil. Plates. Plate I.—Restoration of Agriocherus latifrons. Plate II-VII. —Illustrations of Sphingidz. Plate VIII.—Skeleton of 77tanotherium robustum. Plate [X.—Vertebral column of 77tanotherium robustum. Plate X.—Skeleton of A/etamynodon planifrons. Plate XI.—Skeleton of A/etamynodon planifrons. DATES OF PUBLICATION OF AUTHORS’ SEPARATES. Art. I, March 8, 1895. Art. II, May 18, 1895. Art. III, May 21, 1895. Art. IV, June 17, 1895. Art. V, June 20, 1895. Art. VI, June 29, 1895. Art. VII, Aug. 21, 1895. Art. VIII, Oct. 11, 1895. Art. IX, Oct. 7, 1895. Art. X, Nov. 8, 1895. Art. XI, Nov. 22, 1895. Art. XII, Dec. 23, 1895. Besides the authors’ separates, and in addition to the regular edition of the Bulletin, 100 copies were issued in signatures as printed, each signature bearing at the bottom of the first page the date of publication. Se eVE FEN OF THE AMERICAN MUSEUM OF NATURAL HISTORY. Volume VII, 1895. Article I.—FOSSIL MAMMALS OF THE PUERCO BEDS. COLLECTION OF 1892. By Henry FAIRFIELD OsBoRN and CHARLES EARLE. The archaic basal Eocene fauna, discovered by Cope in 1880 and extensively described since, still presents problems of the greatest difficulty andinterest. What are these peculiar animals? What are their relations to the Mesozoic and Cenozoic mammals ? With the double object of completing the historical series and of further elucidating these problems, the Museum sent an expedi- tion into the arid Puerco region early in the spring of 1892. The expedition was under the able direction of Dr. J. L. Wort- man assisted by Mr. O. A. Peterson, and Mr. Thomas Raffierty in the field. The explorations were in the San Juan Basin of north- western New Mexico. The Museum is indebted to Mr. E. T. Jeffery, President of the Denver and Rio Grande Railroad, for many courtesies. Dr. Wortman gives the following field notes: “The thickness of the beds is roughly estimated at 800 to 1ooo feet, and as far as can be observed they lie conformably upon the Laramie. At no place examined by us can fossils be said to be abundant, but on the contrary most of the exposures are entirely barren. For convenience they are divided into Upper and Lower Beds, but this [ February, 1895.) [1] Tl 2 Bulletin American Museum of Natural History. \Vol. VMI, scarcely gives an adequate idea of the occurrence of the fossils, for the reason that it is only the extreme upper and lower strata that are productive ; the great intermediate part we found to be singularly barren. “The lower fossil-bearing strata occur in two layers, the lower- most of which lies within to or 15 feet of the base of the forma- tion. ‘This is succeeded after an interval of about 30 feet by a second stratum in which fossils are found, and this appeared to be by far the richer of the two. Both of these strata are of red clay, and at no place did we find them more than a few feet in thickness. “The lower horizon we found exposed in two places, viz.: the head of the Coal Creek or Pina Verta Cafion, and some of the upper tributaries of the Chaco Cafion. It is especially and sharply distinguished by the occurrence of the remains of Polymastodon, which appear to be entirely absent from the upper horizon. “Fossils are much more abundant in the upper strata, and wherever a good exposure was found their occurrence could be more confidently looked for. The genera Chzvox and Panto- Jambda appear to belong exclusively to the upper beds. Owing to the widely separated localities and the general scarcity of fossils, it is at present impossible to say whether it is one or several layers that produce the fossils from these upper beds. It is my opinion, however, that there are several layers, and that their vertical range is somewhat greater than that of the lower horizon. The principal localities of the upper strata are as follows: head of Cafon Gallego, Cafion Blanco, Cafion Escavada, and head of Cafion Chaco.” The main systematic determination, and the larger part of the description of this collection is the work of my colleague, Mr. Earle. The Creodont section is entirely his, as well as many original suggestions as to the relationships of the Primates and Condylarths. The following are the principal new featurés : 1. A division of the Eutheria into Mesoplacentalia and Ceno- placentalia, p. 3. 2. A revision of the Classification, Geological distribution, and Phylogeny of the Puerco mammals, pp. 7-10. 1895.] Osborn and Earle, Fossil Mammats of the Puerco. 3 3. Multituberculata: Description of the complete dentition of Polymastodon, pp. 11-15. 4. Primates: Description of the skeleton of J/mzdrodon, with lemuroid characters. The Chriacidz added to this group. The new genus Oxyacodon, pp. 15-23. 5. Creodonta: C/enedon as an ancestor of the Arctocyonide. Description of the skeleton of Dyssacus, an ancestor of the Mesonychide, pp. 26-39. 6. Tillodonta : Description of the skull of Onychodectes, p. 41. 7. Amblypoda: Relationship of Periptychus to this group. De- scription of a complete skull of Pantolambda, p. 43. 8. Condylarthra: Introduction of /zoclenus to this group. Sys- tematic revision of the Periptychide. Hap/oconus is shown to have probably been arboreal in habit. Pyrotogonodon is shown to be related to 7rigonolestes' and the Artiodactyla. We are indebted to Professor Scott for criticisms of the MSS. upon the Creodonta, and to Professor Cope for the loan of type specimens and for assistance in the determination of species. Pade ls OF I—THE PuERCO MAINLY A MegEsozoic FAUNA. The Placentals should be considered as having exhibited two great centres of functional radiation, which were successive and largely independent of each other. The first is represented by the groups discovered by Cope in the Puerco, and now proved to have extended back certainly into the Cretaceous, probably into the older Mesozoic—to these may be applied the term JZesopla- centalia, or Placentals distinctive of the Mesozoic period. The second is the group, the earlier members of which are found in the Puerco, and which developed and radiated in the succeeding Tertiary ; these may be called the Cenoplacentalia, or distinctively Tertiary Placentals. The difference between these two groups 1 Professor Cope has récently substituted this term for Paztolestes, which is preoccupied. 2*A division of the Eutherian Mammals into the Mesoplacentalia and Cenoplacentalia.’ Trans. N, Y. Acad. Sc., June 4, 1804. 4 Bulletin American Museum of Natural History. |Vol. VU, consists mainly in the lower state of evolution and apparent incapacity for higher development exhibited by the Mesoplacen- tals in contrast with the capacity for rapid development shown by the Cenoplacentals. SUCCESSION OF THE NORTH AMERICAN ORDERS OF MAMMALS. Multituberculata’:****""""" S Proto doutes Oacee ee is Triconodonta*****""""" 1: rei Marsupialia Trituberculates............-}- Insectivora. Tillodonta.............4++++ Carnivora. (05 {2Xoyo Koy suit: Mag non en barteabe p-Bed scorns ican @oryphodontal ie] cedeeersl sores pencee ; Proboscidea. Dinocerata.....2.-..---40 ee | Artiodactyla. Condylarthraisr. «2. 4eet ech eser erate | Perissodactyla. Primates:...5......5 sis: Sucot basso: The first of these terms therefore chiefly serves to express the fact that the Mesoplacentals evolved and diverged in North America and undoubtedly in Europe during Mesozoic times in the Jurassic, Cretaceous and Lower Tertiary. Careful studies show that even the Upper Cretaceous mammals had probably already diverged into Ungulate and Unguiculate, Carnivorous and Insectivorous types. ‘This functional divergence reached its climax in the Puerco, which contains several Laramie reptiles, and Cope inclines to consider this epoch as Post-Cretaceous rather than the base of the Eocene. Here the Mesoplacentals display the greatest variety, and are generally characterized by plantigrade feet and tritubercular teeth, for even among the Un- gulata the molar teeth are developed upon the triangular plan, whether bunodont, selenodont or lophodont. We may consider the Middle Eocene Dinocerata, Tillodonta and Creodonta as 1895.| Osborn and Earle, Fosstl Mammals of the Puerco. 5 spurs of this great Mesoplacental radiation, a few of the Creo- donts only persisting into the Mid-tertiary. In opposing Cope’s view’ that this fauna, with the exception of the Multituberculata, is largely ancestral, it is important to em- phasize the fact that we have not as yet connected any of the Mesoplacentals directly by lineal descent with the Cenoplacentals, excepting Euprotogonia, a supposed ancestor of the Perissodac- tyla ; and Protogonodon, a supposed ancestor of the Artiodactyla. A comparison of Tables I and III shows that out of 39 generic and go specific types existing in the Puerco, only & generic types are followed by analogous forms in the Wahsatch, and 3 of these became extinct in the Bridger. But even if more threads of phy- letic descent are traced by future research the fact remains that the great group of Mesoplacentals as such became extinct ; that the first attempt of the mammals at wide functional radiation failed, and that from some comparatively unspecialized spurs of this dying group a new functional radiation began which reached its climax in the Cenoplacentals of the Miocene period, and sub- sequently declined. TaBLE I.—SUCCESSIVE, ANALOGOUS, AND RELATED TYPES. | LARAMIE. PUERCO. | W AHSATCH. Multituberculata..| Ptilodus.............|.- x< | Neoplagiaulax. | Meniscoéssus........- Polymastodon. PAiladonirr inca si75..0-)| CLITOX:. Tillodonta ....... gcics Cute ue iia Psittacotherium. Hemiganus. Conoryctes.....| Esthonyx. Onychodectes. . . | i Creodonta........ red eae ae Clenodon ... . | Anacodon. | ? Batodon tenuis...... WDISSACT Sis .c 2a | Pachyzena. AMES UTIL, AR RL A te er Ste Nh EAR Pantolambda. . ..) Coryphodon. COLLET, ES ERE RE rT eee Euprotogonia...| Hyracotherium.. Protogonodon...| ? Trigonolestes. (? Didelphops) comptus.} Ectoconus. | eee MEMES tee D ofsis\| <°5:5 sya Sy 2S). RSE ee Indrodon ...... | Anaptomorphus. Mixodectes..... Microsyops. 1* Synopsis of the Vertebrate Fauna of the Puerco Series.’ Am. Phil. Soc., Jan. 20, 1888, Pp. 300. * 6 Bulletin American Museum of Natural History. |Vol. VU, The Mesoplacentals cannot be defined as a homogeneous group ; they are very heterogeneous. No attempt is therefore made to define as Cope has defined the Bunctheria (to include the Creodonta, Mesodonta, Insectivora, Tillodonta and ‘Teenio- donta). What chiefly unites the Mesoplacentals is the possession of a large number of very primitive characters, and the apparent incapacity for progressive evolution. The terms ‘inertia’ and ‘ potential,’ although new in paleontology, seem to express most perfectly the cardinal difference between the Mesoplacentals and the Cenoplacentals. The inertia is seen in the inability to shake off the primitive mammalian characters and assume the modern mammalian standard. Wherever they came into competition the Cenoplacentals drove out the surviving Mesoplacental spurs just as the Placentals will in time supersede the Marsupials of Australia. TABLE IJ.—EUTHERIA: PLACENTALIA. MESOPLACENTALIA. CENOPLACENTALIA, Amblypoda (Dinocerata, Coryphodon- (i) Seer tM mo Ns Sane Ae ae Proboscidea. Condylanthirasecvi 5) beieecrr aete bier Diplarthra : Artiodactyla and Perisso- dactyla. Greodonta Ae s.3 sei ae a oR Carnivora. Taillodontar snes o-ckiscieere se eee Rodentia. Inséctivora’ 2. . Sacccsateiao oie ros AWE kemuroideas, 3.3). cas eee eee wa Anthropoidea. Incerte sedis; “Edentata, Sirenia, Cetacea, Pursuing this hypothetical line of division further, an exception to this elimination, by * survival of the fittest,’ is seen perhaps in two great groups still existing which are universally regarded as extremely primitive ; these are the Insectivora and Lemuroidea ; both orders are closely paralleled in structure by a number of Puerco types, although we cannot as yet positively assert that the latter are either true Insectivora or Lemuroidea, It may be that we should regard the Insectivora and Lemuroidea as persistent Mesoplacentals. 1895. | Osborn and Earle, Fossil Mammals of the Puerco. ‘| The division of the Placental orders upon this physiological and developmental basis would, according to this hypothesis, stand as in Table II, in which groups presenting analogous adapta- tions are connected by dotted lines. II.—SyYNOPSIS AND VERTICAL DISTRIBUTION OF THE PuERCO FAUNA. In the following table the classification is to be regarded as provisional, and subject to extensive modification by future dis- coveries. The order Insectivora is probably represented, but by what types is uncertain. The number of forms embraced by the Lemuroidea is also somewhat doubtful. The Amblypoda may grow at the expense of the Condylarthra by the inclusion of the Periptychide. TapLeE III.—Synopsis OF THE PUERCO FAUNA.’ 1. MULTITUBERCULATA. PLAGIAULACIDA., Plagiaulacine. Ptilodus medizevus Cope.-:........ Ps ee ee ee es trovessanhianus Cope. 3.2. «ccs c-eee WNeoplaciaulax americanus, Cope... <2. c1o- < «0: - “aD Polymastodontinz. ObyIMASLOMODtADENSISECOAC aus «00% 5 eye) laps = oh4)| ose < zh SUN ETTERS MGe aol» tte 5 SO peo S Se |e ae x os HIGSIM ELIS GONEN ES soe, te oe hos S515 [tases = aK POH ALIS GOP Ge torte oN ave, daha laxeie elm a= 2 0) [lores = x = [atimrolisvGrpes cts «tere sist ele 3, 22. lara x - SClEMOGUSIO TCI ie ayo hemes a2 0 <= BOLODONTIDA, Chirox molestus Cope.........-. «Sie ec seme PHCALUS, Copeniiey ies spree aie os) 0s tele. <<. 2. PRIMATES. ANAPTOMORPHID. Kadrodon:;malaris Cope. =. 5... =." Set bee OEE oe eee Geers MIXODECTID&. Mirxodectes pungens Cope... 22 ec sem. 2'scee ; GRASSIUSCHILUS: Cup papier ene en haat fess 1 The types not marked with X in the table are those in which no local record has been kept. 8 Bulletin American Museum of Natural History. |Vol. VU, i TasLeE I11.—Continued. Lower. UPPER. CHRIACIDA. Chriacus pelvidens Cope ...-2-0.-2025+- o--s05% a HAbbaLe IRIS CU Paaananacanoasesdvesccac oe palwinil Copan ety -cmierlaien ieee teteletsenen noc es SSO OMAM no Gap doabsocdqcgdoDesoDe Protochriacus priscus) CopZ tra titeteiatste tet etait roi ri Simplex Gopdare eee ee ee See Alana ney Obie IFoo ssh esate coco shoe Epichriacus schlosserianus Cope..........-----.- Incerte sedis. Loxolophus ihyattianus (Copan ne eel errr weeicoX ‘Tricentes bucculentns Cope =r. re teeter olla al ea ie ene. aoe crassicollidense Capea tiene eer Aka Snies Culoierecomis (OG. 5.550500 050005550 Ellipsodon inzequidens Cope ........ -- Be oie 3. CREODONTA. ARCTOCYONIDA, Clzenodon ferox Cope.......-..- a (KopmMUANAS (O7N2coanocs donne soap Gods Proto gonoidesnGofe. aria tin-t elie Tetraclanodon floverianus Cope... TRIISODONTIDA, Triisodon quivirensis Cope ...... cc ae MOA oe biculminatusiGopcme tee rite ene ia Bes) Gua co. heil printanusnGos esate rete iter eet Sarcothraustes antiquus Cofe... ............0.. | edie ne ates ie 1 COLY Patsy COP er ersten arte ett to ee eee CTAaSSICUSPIS Cope yj teehee al ate EPA bathyonathusmGopeo teeter Aout 4 x ae ae ae Goniacodon levisanus Cofe. ae a. © wo" © fe) 6 2S 0, 0) sie. ele, (9 ae) gaudryanusCopes.- «ce eee Bebe rusticus Cope ..... Microcleenodon assurgens Cope MESONYCHID. Dissacus navajovius Cope oe ae ee ed canmifex (Cope We selene PROVIVERRID. Deltatherium fundaminis Cope... 0.2. on osm \ecuuuehrse teers eee MIACID. Didymictis haydenianus Cope rf primus Cope 4. TILLODONTA. Psittacotherium aspasize Cope S'S) e) 8.440 0) 6\% # a's eles slelm multifragum Cope......... zis fait le sola emo) wii] Se eee megalodus Cope Hemiganus vultuosus Cope ae oe eee eet esse ee ee eee tees Otarlidens (Coss hr een sist ee eee aS Ries Conoryctes comma ‘Cape... s GOP Cs. «10% 5 2 15 =. 991-06), Bae, < Anisonchine. PRE CIODOULOLACHIS COPE e tat. iis so 7. INCERTZE SEDIS. Oxyacodon apiculatus Ha7 le sees. 2% 3. 2 +-- Sere Oxyclzenus cuspidatus Cofe..... OSCE ee, Paradoxodon rutimeyeranus Cope................ Fonee 4 Carcinodon flholianus (Cope o-0 pees «1-66 <3 = ESE Mioekenus anterruptis) Cope wees <<. 3 =< -- Sa aybaVban th MO Tse pane OL oe oe Oe x acolytus Cope ..-.... Mea oc, ete A HEM PACOMOMMINVERSUS COPA. eisnsiand o'ciee the) fais 2.2 3 5) s 10 Bulletin American Museum of Natural History. |Vol. VXI, This geological distribution, made up from the field notes of Cope (Baldwin) and Wortman also probably contains many errors of detail, and is subject to alteration by future discovery. The following are the most important points in the vertical distribu- tion : rt. Among the Multituberculata, the Plagiaulacidz including Polymastodon are confined to the lower beds, while the Chiro- gidee are found in the upper beds. 2. Among the Primates three species of Chriacidze occur in the lower beds, while other Primates are found in the upper beds. 3. Among the Creodonta the following are recorded from the upper division only: Clenodon, Triisodon (excepting one species), Dissacus, Deltatherium. Sarcothraustes occurs in both upper and lower. : 4. Among the Tillodonta, Hemiganus and Onychodectes are found in the lower and not in the upper beds. 5. Among the Amblypoda, Pantolaméda is found only in the upper beds. 6. Among the Condylarthra it is remarkable that the highly specialized Fctoconus occurs in the lowermost portion of the lower beds associated with remains of Peripiychus. Periptychus, however, extends also into the upper beds. ‘The specialized Hemithleus is from the lower beds only, while the more simple Hlaploconus and Anisonchus have been found both in the upper and lower beds. Among the Phenacodontide Pyrotogonodon is found only in the lower beds, Zuprotogonia puercencis only in the upper beds. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. II IlI.—SysTEMATIC DESCRIPTION. 1. Order MULTITUBERCULATA Cope. Family PLAGIAULACIDE Marsh. Subfamily POLYMASTODONTIN. The discovery that MWeniscoéssus' of the Laramie is a transition form between Plagiaulax and Polymastodon removes the latter genus to the Plagiaulacidz, subfamily Polymastodontine, char- acterized by reduced fourth premolars. 1. Plagiaulacine. | 2. Polymastodontine. Early representatives of the Plagiau- Latest representatives of the Plagiau- lacidee: Premolars, 4-1. Fourth | lacide: Premolars, 1. Fourth pre- premoiars very large and trenchant. | molar greatly reduced. Genus Polymastodon Cofc. Dentition: 17, Cf, Pi, M$. The enlarged anterior pair of incisors verti- cally striated, enamel wanting on posterior surfaces. Lateral upper incisors (13) small, conic. First upper molars with three rows of tubercles ; second ditto, with two rows and a rudimentary third row. This last survivor of a great Mesozoic order is represented in the collection by remains of 45 individuals, many of which are exceptionally perfect. The five species established by Cope seem to be valid with the exception of P. Zatimolis, which is doubtfully distinct from P. faoénsis. We can now amplify Cope’s definitions as follows : P. foliatus.* P.. taoénsis.* Of small size. Lower molar tuber- Jaws robust. Lower first molar cles flattened, block-shaped. Cusp much larger than second. Cusp for- formula : first molar, 5 outer, 4 inner; mula: first molar, 7-8 outer, 6 inner. second molar, 4 outer, 2inner. P4-— Incisors broad, with enamel band M2=22 mm. _ wanting on fang. 1 See Osborn, ‘ Fossil Mammals of the Upper Cretaceous Beds.’ Bull. Am. Mus. Nat. Hist., Vol. V, 1893, pp. 312-330. y¥ 2 Am. Nat., 1882, p. 416. 3 Am. Nat., 1882, p. 684. 12 Bulletin American Museum of Natural History. \Vol. VU, P. fissidens.' P. latimolis.* Of intermediate size. Molar tuber- Jaws robust. Lower first molar cules conic. Cusp formula: first lower | slightly larger than second. Cusp molar, 5 outer, 4 inner, 3 additional | formula: 1st molar (?), 5 outer, 6 cusplets. M1, 135 mm. inner. P. attenuatus.” P. selenodus, sp. nov. Jaws slender. Molars compressed Laterally compressed. Lower inolar in mid-region. Cusp formula: first | tubercles crescents, opening back- molar, g outer, 6-7 inner; second | wards. Cusp formula: first molar, molar, 5 outer, 4 inner. Incisors | outer 7, inner 6. M1, 2 mm. narrow, with enamel extending to base of fang. Upper incisor grooved laterally. In general P. foliatus is the most primitive type, distinguished by small size and very few tubercles. P. fissédens is somewhat larger, with the same number of full sized conic tubercles, but with accessory tubercles. P. se/enodus is still larger, with more numerous crescentic tubercles; /. attenwatus is laterally com- pressed with long enamel bands on the incisors ; ?. éaoénsis and P. latimolis are robust, with short enamel bands on the incisors. Polymastodon attenuatus Cofc. This species is represented by the dentition of a left mandibu- lar ramus (No. 967) and by a complete upper dentition (No. 970). Also by Nos. 730, 720, 743, 734. The lower teeth correspond in general to the description of Cope;’ the incisor is very long and slender, with well-defined grooves, multiplying towards the fang; the enamel band is confined to half the section. The fourth premolar is narrow, and exhibits three minute apical cusps, the second and third being separated by a deep notch. The first molar is long, narrow, and compressed in the middle ; the second is short and rounded. The complete upper dentition is of great interest. ‘he large incisor ( ? 2) is rather slender, sharply grooved, restricted enamel band and a deep postero-external groove. The lateral incisor (? 3) is a very small conical tooth, compressed antero-posteriorly, with its enamel confined to the anterior surface, probably as an instance of ‘meristic repetition.” The fourth premolar is small, ' Am, Phil. Soc., 1883, p. 322. 3 Am. Nat., 1885, p. 494. 2 Am. Nat., 1885, p. 385. 4 Am. Nat., 1885, p. 404. 1895.] Osborn and Earle, Fossil Mammals of the Puerco. 13 Swe e Ve Fig. 1. Potymastopon. A, P. attenuatus, composition side view of upper and lower jaws. B, Superior dentition, crown view. C, P. tavénsis, lower jaws, crown view. Two-thirds natu- ral size. conical, with two apical cusps. The first molar is long, rather narrower than in /. ¢aoénmsis, and somewhat compressed in the ‘middle ; the second is sub-triangular, narrowing posteriorly, and with only one and one-half cusps in the outer row. (Fig. 1 B.) The molar cusps are transversely oval, with some tendency to 14 Bulletin American Museum of Natural History. [Vol. VII, exhibit crescents opening forwards in the upper series and_back- wards in the lower series, as in eniscoéssus. The cusp formula is: Lower Mo tars. Upper Mo.ars. Outer. Inner. Outer. Middle. Inner. 7 a> are | Hirst molar | nO | 6-7 9 9 10 Second molar... 5 4 I-2 4 4-5 Polymastodon taoensis Cofc. This species includes the robust types, and is represented by numerous specimens—Nos. 742, 746-8, 750, 753, 721-3, 725-32; 735, 736, 743, 968. The lower jaws are robust ; the coronoid rises from the outer side of the third molar and posterior half of the second; the pterygoid fossa is deeply excavated, and the masseteric fossa is a broad concavity ; the lower border of the jaw is thus broad and flat and 1 shaped in section ; the condyle is oval and its long axis is placed obliquely, not antero-posteriorly as in the Rodents. The obliquity is greater in some specimens (No. 734) than in others. When the jaws are spread as in Fig. 1 C, the opposite molars are exactly parallel with each other, and the condyles are transverse, but the antero-posterior grooving of the molars is proof of motion in the same direction. A marked feature of the jaw is that the coronoid rises on the outer side of the second molar. The lower incisors are broader than in /. attenuatus and are readily distinguished by the fact that the enamel terminates at or above the alveolar border, and does not extend down upon the fang. The cusp formula of the molars as compared with Menitscoéssus is : Lower Movars. Uprer Morars. Outer. Inner. Outer. Middle. Inner. | First molar...} 7-9 6-7 9 g-10 10-12 | | Polymastodon. Second molar. 4 4 I 4 5 ( taoénsis, First molar... 5 4 a 7 [eo \ Meniscoéssus Second molar. 4 2 3 4 4 | conguistus. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 15 Although the lower molars exhibit typically but two rows, we occasionally observe (No. 725) a postero-external accessory row upon the first molar, and upon the first and second molars (Nos. 725, 731). The form of the cusps is occasionally subcrescentic. The comparison with Aenrscoéssus shows an average addition of two cusps to the first molars in both jaws, and an apparent degeneration of the outer row in the second upper molar, so that this tooth is relatively simpler in Polymastodon than in the older genus Meniscoéssus. Polymastodon fissidens Coe. This species is represented by a fragment of the left mandible (No. 751), containing the base of P4 and the much worn and fissured crown of the first molar. This tooth is a trifle larger than that in Cope’s type, and the cusp formula is apparently 6 outer, 5 inner; so that there is some doubt as to this specific reference. Polymastodon selenodus, sp. nov. The type (No. 749) lower molar is widely distinct from the above in the crescentic form of its molar cusps. The anterior cusps are distinct, the posterior are low and irregular ; there are 7 in the outer and 6 in the inner row; the anterior border is convex, the posterior is flattened, giving a sub- crescentic section, which reminds us strongly of the cusps of Meniscoéssus conguistus of the Laramie. The fourth premolar is very small. ie Fragments of a Polymastodon skull (No. 734), undetermined, exhibit a broad molar shelf below the orbit. Another skull (No. 721) is still embedded in a very hard matrix. 2. Order PRIMATES. We find in the Puerco numerous remains of the Primates, and there is every reason to believe that these animals were both abundant and highly specialized or modernized. At present, 16 Bulletin American Museum of Natural History. {Vol. VU, however, there is no satisfactory means of determining as regards several of these types whether they belong to the Lemuroid or to the Anthropoid phylum; we refer especially to Zr7centes Cope, to the related /zdrodon Cope, and to AZixodectes. Of Lndrodon we have the first remains of the skeleton which have been found in the Puerco, by far the oldest Primate skeleton known. Appar- ently related in dentition to modern Lemurs are the Chrzacida, a family including larger forms which we remove from the Creo- donts where they have been placed by Cope, and provisionally refer to the Primates. 1. Chriacide. 2. Anaptomorphide. 3. Mixodectide. Incisors normal. Pre- Incisors normal, $. A pair of incisors en- molars, 4. Premolars | Premolars, 3-2. larged. Premolars, 3. spaced. Pmq4 without tritocone. Molars tri- tubercular. Family ANAPTOMORPHID Cofe. Genus Indrodon Cofe.' (?) Dentition : Ii C1, P?, M#®. Premolars spaced and conic. Upper molars with flattened outer cusps, a rudimentary postero-internal tubercle or hypocone. This type is distinguished from Anaptomorphus by absence of internal lobe upon third superior premolar, and by spacing of premolars. Indrodon malaris Coc. In Cope’s type, a fragmentary skull, the maxillo-premaxillary suture cannot be made out; the homologies of the anterior teeth are therefore uncertain ; they apparently represent two incisors, and a canine. The second and third upper premolars are small, conic and widely spaced ; the fourth premolar only has a conic internal lobe. ‘The true upper molars are low-crowned and sub- triangular ; the outer cusps are flattened and there is a wide external cingulum, marked by minute cingules. The intermediate tubercles are absent or worn off in the type; there is also a faint postero-internal cingulum, and the hypocene is represented as a ! Proc. Am. Phil. Soc., 1883, p. 318. ae ee Oe 1895.] Osborn and Earle, Fossil Mammals of the Puerco. 17 cingular cusp upon the first and second molars. The third molar is somewhat smaller than the others, and there is no such great inequality in size as we observe in 77icentes bucculentus, or in T. inequidens. This description refers to Cope’s type. SKELETON OF INDRODON.—The reference to /udrodon, of the skeleton No. 823, is somewhat doubtful, because the upper molars associated with the skeleton are so much worn. The material consists of fragments from all parts of the skele- ton, which were collected by Dr. Wortman withthe greatest care, including: Teeth, superior P4—M3 inclusive ; inferior P3 and M1; part of the lower jaw, and isolated M1, and P3. Of the vertebral column are pre- served: cervicals, 2; dorsals, Fig. 2. Indrodon malaris. Superior molars, 6; lumbars, 4; sacrals, 1; "4 an inferior true molar. “Twice natural size caudals, 7; these are mostly centra with portions of the neural arches. Of the appendages, portions of the scapula, humerus, radius and ulna, metacarpals and phalanges are preserved. Of the hind limb parts of the ilium, femur, tibia and fibula and tarsals are preserved. The animal (No. 823) was about half the size of Lemur varius, with slender limbs and a long powerful tail; in fact it closely resembled some of the living Lemurs. The principal characters are as follows: Dentition.—Vhe fourth upper premo- lar has a sharp prominent external pro- tocone and an internal deuterocone, with the rudiment of the tritocone. The molar crowns, although broken, in- dicate that they were tritubercular, wider transversely, and more com- pactly placed than in the Zudrodon type, although of the same measurements. The intermediate tubercles are in- tions of the skeleton: A, Proximal portion ofrighttibiaand fibula: B, Gistinguishable, owing to excessive Head of humerus; C, Left tarsus. wear. Natural size. Fig. 3. Zudrodon malaris. Por- [ February, 1895.| 2 18 Bulletin American Museum of Natural History. [Val. VII, The posterior portion of the lower jaw contains the worn crowns of the first and second molars. Fore Limb.—The fore limb characters are the following: Scapula, with an obtuse coracoid; humerus, with tuberosities not very prominent, but exhibiting a marked deltoid ridge extending from the greater tuberosity on the outer side of the front face of the shaft, and a marked ridge extending from the lesser tuberosity down the inner side. A similar relation of these ridges is very characteristic of the Lemurs, and is also seen in some of the true Monkeys. In most of the Monkeys, however, the deltoid ridge occupies a median position on the front face of the shaft, and the lesser ridge is reduced or wanting. Distally the humerus presents a strong entepicondylar foramen. The head of the radius is oval, and the ulna has a short olecranon. Hind Limb.—The ilium has an imperforate acetabulum and a wide cotyloid notch. The femur exhibits three trochanters, the third trochanter extending about half-way down the outer side of the shaft ; the head exhibits a pit with a round ligament ; the cnemial crest of the tibia is prolonged down the front face of the shaft. The fibula is well developed. ‘The astragalus exhibits the astragalar foramen, and a large fibula facet, similar to that in Lemur varius ; it has a deep posterior groove for the flexor tendon ; distally the astragalus has a short neck and a convex navicular facet. The cuboid is subquadrate in form. Unlike the Condy- larthra articulation, the calcaneo-cuboidal facet is nearly flat. Vertebre.—The axis exhibits a short odontoid process ; the cer- vical centra are flat. The dorsal centra are triangular in form ; the lumbar centra are more elongate and flattened. ‘he detached zygapophyses which belong in the posterior dorsal or lumbar region, exhibit convex vertically placed facets. There is appar- ently but a single sacral vertebra. The caudals are long and well developed. ? Upper DENTITION oF INDRODON.—Portions (No. 833) of the maxillee and of the lower jaw with certain teeth, were found with remains of two skeletons, a larger (No. 833) and a_ smaller (No. 834). 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 19 The identification of No. 833 with /zdrodon is not absolute ; it is based upon the likeness of the superior molars (Fig. 4) to the somewhat fractured crowns in Cope’s type. It is rendered doubtful by the compactly placed lower premolars. The fourth upper premolar is triangular, with a complete investing cingulum, a high protocone, a deutero- cone, and incipient tritocone. The molars pee Ef hemi onc are beautifully preserved ; they consist of oT Ee Ne a perfect trigon with a detached spur-like hypocone upon the first and second molars ; the external cusps are subcrescentic, and in the sharp external cingulum we observe a distinct mesostyle and less prominent para- and metastyles ; the intermediate tubercles are developed upon the spurs between the external and internal cusps. The jaw contains the alveoli of a ~ small lateral incisor, a larger canine, a small one-rooted premolar ; next a two-rooted premolar, followed by a premolar crown which A is either P3 or P4; this has a sharp crown and a low heel. If this is the fourth premolar it is closely similar to that of Anaptomorphus. A comparison with Axaptomor- phus homunculus’ of the Wahsatch shows a very similar configuration of the lower jaw, and probably a ee similar lower formula, re M 7 jaw, internal and upper view. Natural The upper teeth of this specimen cm (No. 833) differ from those of A. homunculus in the much more prominent hypocone spur. The humerus associated with this specimen also has the double ridge observed in No. 823. 3 2 .- i Incerte sedis.—The skeleton (No. 834) found with this type is of smaller size and presents many differences from that of No. 823. Lower Jaw.—Fig. 5 represents a lower jaw (No. 829), which we provisionally refer to this genus owing to the similar dimen- sions of the lower molar series with those in the fractured jaw 1 Bull. Am. Mus. Nat. Hist., 1892, p. 103. 20 Bulletin American Museum of Natural History. |Vol. VU, attached to Cope’s type skull. The Pmgq is a rounded cone with a faint deuteroconid and anterior basal cusp; the talonid is broad and basin-shaped. ‘The molars are distinguished by the absence or vestigial condition of the paraconid, the elevation of the trigo- nid, the rather depressed but distinct hypoconid from which extends inwards a broad internal basin representing the fusion of the hypoconulid and entoconid. In M3 the hypoconulid is a distinct cusp. An isolated single molar (No. 829a) shows the same characters. Family CHRIACIDA, fam. nov. This family includes forms more primitive than the Adapide but with a similar dental formula. It is exceedingly difficult in the present state of our knowledge to decide with certainty as to the ordinal affinities of the genera which Scott’ has included in the family Oxyclaenide. We think it probable, however, that Ciriacus and its allies are more closely related to the Primates than to any of the Creodonta to which Cope has referred them. The type species of the genus Chrvacus, namely, C. (Pelycodus) peloidens, was in fact at first included by Cope with the Lemurine like Pe/ycodus. Scott has suggested the Primate relationship of these forms. Schlosser has also observed the resemblance in the shape of the jaw in Chriacus to that of Necrolemur. We here discuss the evidences of Primate relationship. Of the three points spoken of by Scott as separating Chriacus from the Lemurs, namely, (1) the character of the jaw symphysis, (2) spacing of the inferior premolars, (3) the presence of the para- conid; the second character at least occurs. in the Bridger genus Tomithertum, which is an undoubted Primate. Again, among the recent Lemurs, the last two inferior premolars are spaced in some species of Zemur, while the paraconid is present in Zarsius. Thus the difficulties raised by Scott are all removed. As remarked by Scott the superior molars of Chréacus are sur- prisingly like those of certain Lemurs, and it is to be emphasized ' These genera employed by Scott upon types of Cope’s species of Mzoclenus, Chriacus, and Tricentes areas follows: Oxyclenus, Chriacus, Protochriacus, Epichriacus, Pentacodon, Loxolophus, Tricentes, Ellipsodon. 1895.] Osborn and Earle, Fossil Mammals of the Puerco. 21 —— — that they resemble those of the Lemurs more closely than those of the Creodonts. Again, we have some remains of the skeleton of a form, probably belonging to the Chriacide, which resembles corresponding parts in the recent Lemurs. There are apparently two types of mandibular symphysis among the primitive Primates of the Puerco; in Chrtacus this portion of the jaw is long and narrow. ‘This is the most primitive condition, and common to many Puerco forms. The other type (Anaptomorphide, Mixo- dectidz) presents the jaw at the symphysis as deep as below the last lower molar. Genus Chriacus Cope. Pelycodus Cope, in part. Dentition : P?, M3. Superior true molars tritubercular with hypocone, and on second molar an antero-internal cusp (protostyle); third upper molar reduced in size. First and second inferior premolars spaced, fourth with a deuteroconid and heel. Trigonid of inferior molars higher than talon ; paraconid present. The superior molars in this genus especially resemble those of the genera Lemur and Galago. As compared with the upper molars of such a typical Creodont as De/tatherium those of Chria- cus differ (t) in their more square form ; (2) in the rounded shape of the external cusp ; (3) in the large development of the supple- mentary internal cusps. The lower true molars of Chrzacus have the trigonid less elevated than in De/tatherium, and the talon, similar in shape to that of Pelycodus, resembles that of the Lemurs in being very broad and wide. Chriacus baldwini Cope. There are only portions of jaws of this species in the col- lection (Nos. 789, 811 and 812). In the C. da/dwini the first pre- molar of the lower jaw is separated by a wide interval from the second ; the second and third are nearer together, and there is no interval between the third and fourth. The crown of the fourth premolar is high, recurved, and much elevated above that of the first true molar. The mandible is elongated and becomes slender in the premolar region ; its general form closely resembles that of Protochriacus priscus. 22 Bulletin American Museum of Natural History. |Vol. VU, Genus Protochriacus Scoé¢. Superior molars with no protostyle, and hypocone very weakly developed. Inferior true molars with trigonid little raised above talon. (Type, ?. prisczs.) This genus slightly differs from CAriacus in the more primitive structure of its upper true molars, which have the supplementary antero-internal cusp hardly developed at all. ‘The inferior true molars differ widely from those of Cirzacus, and these teeth in the two species included by Scott in Protochriacus are quite different in structure, and further investigation may prove that they belong to distinct genera. Protochriacus priscus Cofe. Inferior true molars with no external cingulum ; trigonid only slightly raised above talon. Talon of last lower molar very wide and deep, with hypoconulid. The material in our collection pertaining to this small species is very abundant; there are no less than a dozen fragments of jaws and upper teeth (Nos. 802, 803, 817, 818, 939). The first and second inferior premolars are spaced, and the last premolar has no deuteroconid. The paraconid is small and placed between the proto- and metaconids, but nearer the latter cusp. The man- dible is long, narrow, and tapers gradually to the symphysis, which is much narrower than the portion below the true molars. The inferior true molars of this species are more of the Lemur type than those of the allied species, viz., P. s¢mplex. ‘The talonid is wide and is more extended transversely than the trigonid ; the cusps forming the border of the basin-like talon are not distinctly separated from each other as in P. semplex. Protochriacus attenuatus, sp. nov. Paraconid well marked, on a line with metaconid, trigonid not raised above talon, hypoconulid distinct. The type of this new species of Protochriacus is specimen No. 790. This specimen is smaller than the P. priscus; the jaw is very narrow and slender. The crescents of the inferior true molars are strongly marked, and the cuspsare sharper than in the 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 23 allied species. The paraconid is well marked on the first true molar, but is rudimentary on the second. The shape of the talon of the last tooth of this series is quite different from that of P?. priscus. Measurements. oralgength of Miranda: (20 a... 5.16 -.2 .O12 Wepthorgaw below ME... sh. sce ceacsss ces .008 Protochriacus simplex Co/e. Inferior true molars with trigonid much raised above talon, paraconid well developed. A strong external cingulum on all the lower true molars. Talonid of last lower molar much smaller than trigonid, with hypoconulid well con- stricted off. . The type of lower molar found in this species is more like that of the typical Creodonts (De/tatherium); the trigonid is high and the anterior portion of the same is more thrown out than in P. priscus, thus giving the teeth a more trenchant function. One specimen (No. 799), among others, in the collection of this species, has the upper molars associated with the lower teeth. The superior molars are much extended transversely, more so than in P. priscus ; the external cusps are round in section, and the postero-internal cingulum is not as much developed into a hypocone as in ?. priscus. In specimen No. 793 the jaw is much deeper than in No. 794, however the teeth are nearly of the same size ; great variation in the depth of the jaw is often displayed by the same species of Puerco mammals. Genera INCERTX SEDIS. Genus Tricentes Cope.’ Dentition : I?, Ci}, P3, M2. Premolars spaced and conic. Molars with rounded tubercles, hypocone well developed. Molars irregular in size; third molar reduced. ‘Trigonid slightly elevated. Paraconid reduced. 1 Proc, Am. Phil. Soc., 1883, p. 315- 24 Bulletin American Museum of Natural History. |Vol. VU, Tricentes bucculentus Cope. The third upper premolar triangular ; the fourth with a compressed protocone and a large internal cusp; first and second upper molars with hypocone, third small, tritubercular ; lower molars with tubercular talonid. There is only one example (No. 784) of this species in the collection ; this includes both upper and lower sets of teeth, and as the latter have not been described, this specimen is of importance. The roots of the upper canines of both sides are preserved, showing that these teeth were quite long and powerful; they are separated from the second premolar by a wide diastema. The first premolar has disappeared. The crowns of the second and third premolars are broken off, the fourth premolar has a high protocone and a well-marked deuterocone. A very minute cin- gular hypocone is present on the first molar, but on the second molar the cingulum is not so distinctly developed into a hypo- cone. The last upper molar is small and has two external cones. The great size of M2 as compared with M1 and M3 1s to be noted in this species. The lower jaw contains the crowns of the third premolar and the second and third molars. The crown of Pm3 is very slender, without a heel; there is some indication that there was a minute second premolar in front of this tooth. The second lower molar has the trigonid slightly raised above the talon, which has a dif- ferent form from that of Prolochriacus; the paraconid is present, but greatly reduced. In 7. éducculentus the talon is notched at its posterior border by a posterior cingulum which extends to the slightly developed external cingulum. The last lower true molar has an elongated talon, as in the Bridger Monkeys ; this extends postero-internally into a high ridge upon which the entoconid is not differentiated, although the hypoconulid is well marked. The lower true molars of 7Zvicentes remind one strongly of those of M. turgidus. The jaw is long and deep beneath Pmz1 as it is beneath the last true molar. Incerte sedis.—A lower jaw (No. 815) containing an incisor, canine and two molars is provisionally placed here. ‘The incisor is small and spathulate. The molar tubercles are all upon the same level, the trigonid not being elevated. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 25 Genus Oxyacodon, gen. nov. Fourth lower premolar strongly compressed laterally, with only a very minute talon, no deuteroconid. Crowns of inferior true molars high with sharp cusps, trigonid not elevated above talonid, paraconid reduced. Hypoconulid of last lower molar high and sharp. The type of this new genus is a fragment of a lower jaw bear- ing the last lower premolar and first two molars (No. 816). There is also another portion of a jaw which we refer to this genus (No. 806). The true molars in this genus resemble somewhat those of Anzsonchus, but the structure of the last premolar is widely different. The general structure of the teeth differs decid- edly from that seen in Chriacus or Tricentes, and appears to be more of the insectivorous type. Oxyacodon apiculatus, sp. nov. Last lower premolar higher than the first true molar, and the crown of same as long antero-posteriorly as the latter. Hypoconulid of M3 well constricted off ; very sharp and curved forwards. The last lower premolar is flattened with sharp anterior and posterior cutting edges ; there is only a very slight enlargement behind. ‘This tooth differs from that of Protochriacus in being more flattened and trenchant. The second true molar is high and narrow with four principal cusps inclined forward ; these cusps are also less connected than in the typical genera of the Chriacidz. The structure of the Fig. 6. Oxyacodon apiculatus. talon of the last lower true molar is Portion of left lower jaw, external : 5 re i view. Composition. Oneandone- — peculiar, in arising from the height of half natural size. 3 : a the hypoconulid, which is unusually sharp and pointed. ‘The jaw is deep, and was probably short. This character relates this genus to the Primates. ? Chriacus ——— A jaw (No. 835), with fragmentary remains of a skeleton, is of importance. ‘The jaw is not at all like that of the recent Lemurs, 26 Bulletin American Museum of Natural History. [Vol. VU, but resembles in form that of Chrzacus and its allies ; that is to say, the symphysial part is much elongated, slender and slopes gradually to the symphysis, instead of being deep and abrupt as in the recent Lemuroidea and Anthropoidea. Unfortunately the teeth are all absent from this specimen, so that we cannot identify it with certainty. The alveoli of the premolars are quite distinct, the first is piaced close to the canine, the second is spaced as in Chriacus, the last three premolars are two-rooted. Measurements of Jaw. M. Length of inferior molar series ...... ie Seba .040 Depthory awe ats itae rd eetetes setter eee > "%OLO si ie VSAM Ss oboe coteonas oe .008 The part of the humerus associated with this jaw is extremely long, and it is of interest to note that the two proximal crests’ so characteristic of recent Lemurs are present on this specimen. 3. Order CREODONTA Cope. Family ARCTOCYONIDE Cofe. Genus Clenodon Sco/z. Mioclenus COPE, in part. Superior molars subquadrate in outline, with well-developed hypocone on the first and second. Inferior premolars simple in structure, last without meta- conid. Inferior true molars with trigonid on a level with talonid, and cusps of same not distinctly differentiated. Borders of molars and edges of premolars serrated. This genus is easily distinguished from other Creodonta of the Puerco by its low-crowned molars, in which the cusps are little raised above the general surface of the teeth. The crowns of the lower premolars are sharp and high, and the last tooth of this series 1s without a heel. C/enodon, as shown by Scott, is closely related to the European genus Arcfocyon. 1895.] Osborn and Earle, Fossil Mammals of the Puerco. 27 Clzenodon ferox Cope. Crown of last inferior premolar much higher than that of first true molar, and provided with a well-marked external cingulum. Second and third inferior premolars much reduced in size. Crowns of lower true molars very flat, with cusps hardly distinguishable ; hypoconulid of M3 large and covered with crenu- lations. There is only one specimen of this rather rare form in the col- lection (No. 772) ; this is a jaw in which the last three premolars and the true molars are well preserved. The first two premolars are much smaller than the last tooth of this series ; the crowns are rather high and compressed. The fourth inferior premolar has a high crown which is recurved. A peculiarity of the pre- molars is the serration of their anterior and posterior edges, as in the Reptilia. This serration can be plainly felt in running the finger over the edges of the teeth, although not well marked to the naked eye. The structure of the crowns of the lower true molars reminds one strongly of those of the Wahsatch Anacodon ; the borders of these teeth are slightly raised above the general surface, but not produced into well-marked cusps. The last lower true molar is unusually flat and ill defined in the structure of the crown; it has five slight elevations corresponding to the cusps of more highly developed forms, and the enamel surrounding the cusps is much crenulated, like that of Anacodon. The hypoconulid is peculiar in being very flat and much extended posteriorly. It is interesting to be able to trace out another line of descent from a Puerco to a Wahsatch form, and we think it certain that Clenodon is the ancestor of the peculiar Wahsatch type Anacodon. Only recently Osborn and Wortman’ have removed Axacodon from the Condylarthra and placed it in its true position near Arctocyon. This is indicated not only by the structure of the molars, but by the incipient atrophy of the premolars. The anterior lower premolars of C/enodon are very small and are undergoing a rapid reduction in their size; the first lower premolar is still present in C/enodon, but absent in Anacodon. ‘The lower true molars in these genera resemble each other very closely in struc- ture, and in both the crowns are much flattened and covered with prominent crenulations of the enamel. 1 Bull. Am. Mus, Nat. Hist., 1892, p. 115. 28 = Bulletin American Museum of Natural History. (Vol. VU, Family TRIISODONTID4E Sco/z. Genus Triisodon Cofe. Triisodon biculminatus Cope. A fragment of a lower jaw (No. 774), with the true molars intact, is in the collection, and probably belongs to this species. As in Cope’s type specimen, the talonid is largely developed, and Fig. 7. Zyitsodon biculminatus, Internal and crown view of inferior molars. Natural size. the entoconid is not distinctly separated from the hypoconulid. The paraconid of the first true molar is submedian in position and well separated from the cusps behind ; on the second and third molars this cusp is only weakly developed. ‘The hypoconu- lid of the last inferior true molar is semicircular in form, convex posteriorly, and concave anteriorly. ‘Traces of the talonid on the last inferior premolar are preserved, showing this to have been much extended behind as in 77¢sodon quivirensts. Genus Sarcothraustes Co/c. Mioclenus Core, in part. Dentition : Ij, C1, P3, M3. Superior true molars with paracone and metacone conical and equal in size. Last superior premolar not molariform, and same tooth of the lower series with talonid consisting of two cusps. Inferior true molars with trigonid raised above the talonid, the former consisting of three cusps with the protoconid much larger than the para- or metaconid. Metaconid distinctly separated from the protoconid and on the same fore and aft line with the paraconid. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 29 Individuals of the genus Sarcothraustes are the most numerous of any of the Puerco Creodonts. This genus is very closely related to Z7z7sodon, and is difficult to separate generically from Goniacodon. Sarcothraustes antiquus Cofc. This species is represented in the collection by a single jaw (No. 785). This specimen has only three teeth preserved, the two posterior being true molars, but they are not well enough preserved for a detailed description. However, from their worn surface, we think it probable that they resembled in structure those of .S. corypheus. The tooth which Cope in his type speci- men identified as the first true molar, is really the last premolar; and in the American Museum specimen there are the alveoli for three premolars anterior to the latter, thus proving that Sarco- thraustes had a full complement of premolars below. The last true molar is absent in this specimen. Sarcothraustes coryphzus Cofe. Numerous remains of this species are to be found in the col- lection, it being represented by more specimens than any other Creodont. Of these, Nos. 764, 762, 765 and 766, are the best preserved. In No. 764 fragments of the skull with the greater part of the upper dentition are present. Dentition.—The canine is small and weak in this species, and diverges considerably from the palate ; behind this tooth there are alveoli for two premolars, the last upper premolar being well preserved. This proves conclusively that this genus has only six upper teeth behind the canine, or the same number as in the Bridger MWesonyx. It differs from AZesonyx in the fact that the last upper premolar is not molariform. ‘The last upper true molar in our specimen has two external cusps. ‘The metacone is smaller than the paracone. Specimen No. 762 presents both upper and lower teeth from the same individual; the lower jaw of this speci- men belongs to the S. dathygnathus type. This is a remarkable jaw, owing to the small size of the teeth and the great depth and length of the mandible. ‘The angular portion of the jaw is much 30 Bulletin American Museum of Natural History. [Vol. VII, extended beyond the teeth and is very heavy. The superior molars found with this jaw are only slightly larger than those of S. corypheus, and we believe accordingly that the specific charac- ter upon which the |S. dathygnathus has been proposed by Cope is merely a case of individual variation. In fact, another jaw in the collection (No. 765), is intermediate in dimensions between the typical S. corypheus and 5S. bathygnathus. ‘Vhe lower teeth of S. corypheus are easily distinguished from those of Dissacus carnifex by their tuberculated talons, which in the latter form are tren- chant. The presence of the hypoconulid on the last lower molar is a marked character of the genus Sarcothraustes. A portion of a cranium exhibits a much elongated, thin and high sagittal crest. The postglenoid process is more extended transversely than in /e/’s, and resembles more in form that of the carnivorous Marsupials. The lower half of a humerus was found with this specimen, and may belong to the same individual. As compared with the size of the skull, it is very small and weak. The deltoid crest is high and extends far down on the shaft. An entepicondylar foramen is present, and the radial trochlea is much extended and slightly convex. Family MESONYCHIDE Cofe. Genus Dissacus Cofe. Dentition: I‘, Ct, P4, M%. Superior true molars with metacone much smaller than paracone. Last upper molar much reduced in size. Lower true molars with protoconid larger than anterior basal tubercle paraconid, and on the same straight line with it. _ Metaconid present on second inferior true molar, and may be absent on the first and also on the last tooth of this series. Dissacus carnifex Cope. This species is represented in the American Museum collection by portions of two skeletons, Nos. 777 and 776. ‘The most com- plete specimen, No. 777, consists of the lower teeth with parts of the skeleton, including a nearly complete carpus and some of the tarsal bones. ‘The importance of this specimen will be appre- ciated when it is known that it is the most complete skeleton of a Creodont ever discovered in the Puerco Beds. 1895.] Osborn and Earle, Fossil Mammals of the Puerco. 31 Dentition—The upper dentition will be described from speci- men No. 776. In this example the teeth are not attached to the maxillary bone, and it is with some difficulty that we are enabled to place them in their proper relation to each other. It is quite certain, however, that with the possible exception of the first upper premolar, all the teeth to be described are properly iden- Fig. 8. Dzssacus carnifex. Crown view of superior, and internal view of inferior molars. Natural size. tified. ‘The supposed first upper premolar is somewhat elongated from before backwards, consisting of a protoconid and a small talon. In the second premolar the principal cone is somewhat triangular in shape, with the heel placed at the postero-external border of the tooth. The third premolar has a small internal cone, with also an antero-external basal cusp. The fourth pre- molar has a tritocone, and also a postero-intermediate tubercle. The upper true molars of Déssacus carnifex resemble closely those of Pachyena ossifraga. This is shown in their much enlarged paracones and the small size of the metacone. In the last upper molar of D. carnifex the metacone is rudimentary, and the whole tooth is much reduced in size. The third lower pre- molar exhibits no anterior basal cusp, a character in which our 322 Bulletin American Museum of Natural History. [ Vol. VII, specimen differs from the type of D. carnifecx of Cope.’ How- ever, in D. navajovius, the smaller species of this genus, the third lower premolar is without an anterior basal cusp. The total length of the lower teeth in the American Museum specimen is greater than in Cope’s type of D. carnifex. We believe that these different characters are individual variations of the same species, and cannot be treated as of specific value. Inthe Ameri- can Museum specimen of D. carnifex both the last two lower true molars have well-developed metaconids, but in Cope’s speci- men of this species the metaconid is absent on the last molar. Skeleton.—The distal portion of a humerus is preserved; this is very broad and heavy, with a prominent entepicondyle. The internal flange of the humerus is strongly marked, and the exter- nal trochlea for the radius is convex and prominent. ‘The proxi- mal end of the radius is much extended transversely, and below this portion the shaft is flattened, as in the plantigrade Carnivora. The bicipital tubercle of the radius is elongated, and not as prominent as in recent forms. ‘The distal articular surface is very heavy and thick from before backwards. ‘The articular surfaces for the scaphoid and lunar are well marked, but not separated by a ridge. The radial styloid process is only slightly developed, and not elongated, thus differing from such a plantigrade as Ursus. The ulna has a heavy, elongated olecranon, and the diameter of its shaft much exceeds that of the radius. The rela- tive proportionate widths, taken by the bones of the forearm in their articulation with the humerus, show that the radius spread over about two-thirds of the width of the humeral trochlear sur- face, thus largely excluding the ulna from articulating anteriorly with the humerus. The radio-humeral articulation in Drssacus is greater than in the Bears, and this denotes less power of supi- nation than in the latter form. Manus,—Vhe general characters of the manus are very primi- tive, but in some respects—as for example, in the displacement of the metacarpals upon the podial elements—a considerable mod- ernization has taken place. As compared with the manus of other Creodonts described hitherto, we find that of Déssacus closely ' See Tertiary Vertebrata, Plate xxv, Fig. 1. 1895.] Osborn and Earle, Fossil Mammals of the Puerco. 33 resembles in its stage of displacement and form of its carpal elements that of A/Zesenyx, as figured by Scott,’ although D. car- nifex has not lost the first digit as in AZesonyx. The scaphoid is exceedingly flat and elongated transversely, with the internal border much thicker than the external, the superior facet is Fig. 9. Dissacus carnifex. Right manus, anterior view. Two-thirds natural size. convex and its backward extension is limited; we cannot with certainty define the facets of the lower surface of the scaphoid ; the surface next to the lunar was occupied by a large centrale, although this bone is unfortunately lost. The position of the centrale was like that in the manus of Wesonyx, namely, above the magnum and to the internal side of the lunar. We would add that this position of the centrale appears to be typical of the 1 Jour. Acad. Nat. Sci. Phila., Vol. PX, Pl. vii, Fig. 1. [ March, 1895.) o family Mesonychidz, whereas in the Hyznodontidz the centrale is interposed between the magnum, scaphoid and lunar, but it is placed partially beneath the latter bone. The lunar is convex above, the posterior concave surface sloping abruptly from the anterior portion ; the posterior hook is not prolonged lower down than the apex of the anterior surface; the inferior articular sur- face is divided nearly equally; that for the unciform, however, 1s slightly larger than the facet for the magnum ; the lunar-centrale facet is triangular in form and placed on the internal face of the bone; the shape of the cuneiform is strikingly like that of Phena- codus; it is much elongated transversely, and has only little depth; the articular cavity for the ulna is deeply concave and only takes up a small portion of the superior surface ; the cunel- form-pisiform facet is very large and oblique to that for the ulna. The pisiform is broad proximally and placed horizontally, upon the cuneiform, like that of the Bears. The shape of the unciform is closely like that of MZesonyx,; it is much extended transversely, and presents a large facet for the lunar ; the internal face has an elongated facet for the third metacarpal, and the inferior surface is flattened and undivided. The unciform in Drssacus exhibits no posterior decurved process as in the Perissodactyl Ungulates. The characters of the magnum are of interest, and its relations to the other bones of the carpus are like those of Wesonyx ; the proximal facets are divided by a slight ridge, but there is no concavity upon the upper surface as in the magnum of the Felide and in that of the Hyendontide. The magnum-lunar facet is broader than that for the centrale ; the posterior convexity of the magnum rises only slightly above the plane of the anterior portion of the superior surface ; the form of this carpal is de- pressed and much extended transversely ; internally it shows a broad and continuous facet for Mc. II. Owing to the large contact between the Mc. III and the unciform in this type, the magnum is placed higher up in the podium than in “Zyenodon, and results in crowding out the centrale from the under surface of the lunar. The trapezoid is triangular and de- pressed ; the trapezium is wanting in this specimen. ‘The proxi- mal portion of the first metacarpal is present, and proves that this digit was of considerable length. ‘Phe second digit is short, 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 35 heavy and proximally articulates by a broad facet with the mag- num. The third metacarpal has a large articular surface for the unciform, and the magnum facet is obliquely placed. The fourth metacarpal is nearly as long as the third; its proximal facet is flat and horizontal in position. The outer metacarpal is short, stout and proximally, on the external side, exhibits a prominent tubercle. The phalanges are much elongated, as is the case generally in plantigrade forms. The ungual phalanges are like those of Mesonyx, being strongly depressed and split at the end. In the Hyznodontide they are shorter and more curved than in the Mesonychide. Pelvis.—One os innominate bone of the left side is in a fair state of preservation ; this shows that the pelvis was much elon- gated, as in AZesonyx. The section of the base of the ilium is tri- angular, with a very prominent ‘rectus’ tubercle. Between the acetabulum and the distal expanded portion of the ilium there isa contracted neck, which is narrow. The acetabular cavity is large and is bordered above by an oblique and broad plate of bone. As compared with the pelvis of Fe/7s and Ursus, we observe that the ischial segment in D7ssacus rather resembles that of the Cats; this is shown in the broad descending processes of the ischia and in the prominent tuberosities of the same. A marked feature of this pelvis is the great elongation of the ischial portion as com- pared with the iliac segment, and we believe this to be a primitive character, for in modern Carnivora the ischial part of the pelvis is much shorter than the iliac. We may add that in the Ungulata, and especially in the more swift-footed members of the same, such as the Artiodactyla, the anterior and posterior divisions of the pelvis are more nearly of the same length than in the Car- nivora., Hind Limb.—The femur is long and its shaft is flattened trans- versely, this widening of the shaft being in strong contrast to the rounded femora of recent Carnivora; the third trochanter is prominent and situated at about one-third the length of the shaft below the great trochanter ; the distal articular surface is heavy, and the shaft is much expanded above the condyles. The length 36 Bulletin American Museum of Natural History. (Vol. V1I, of the femur as compared with the tibia is much greater, and they bear the same linear relations to each other as in the planti-. grade Carnivora (Ursus). The proportions of these bones, and the characters of the manus and pes clearly prove that Dessacus was a semiplantigrade. In Scott’s restoration of AZesonyx the length of the femur is equal to that of the tibia, and as MMesonyx was a digitigrade Creodont, we believe these proportions to be correct. The crest of the tibia in D. carnifex is not raised and elongated; in this character this bone resembles that of the Bear ; the distal articular surface for the astragalus is nearly plane, although there is a slight median convexity and a faint lateral concavity on each side of the latter; the internal malleolus is broad and much prolonged beyond the articular face. Measurements of femur No. 777. IMG 5 5-caeshoodsosduas eSdonoadbonoUN .220 Bread thy LrAnS as PLOKe mientras ie tetel teeter .070 ce LS pa LIS Gyo auch nals Hei asa aS oe ae .054 Pes.—The calcaneum is rather long and slender ; the calcaneal tuberosity is elongated and much compressed, its form more like that of the digitigrade Creodonta ; the ectal facet is placed high above the sustentaculum, being round and not prolonged forward as in the Bear. The transverse diameter of calcaneo-cuboid facet is greater than the vertical, and these relations are the same as those of the Bear. In the digitigrade Carnivora, on the other hand, this facet is nearly round. The astragalus is depressed and broad; the trochlear surface is only slightly concave, and is bor- dered posteriorly by a large foramen, which is of such constant occurrence in Puerco mammals; the large flange-like process bordering the ectal facet is very prominent in this astragalus ; the neck is long and slender, the trochlear surface extending far for- ward upon it ; this extension of the articular surface of the astra- galus has been also pointed out by Cope; the navicular face is convex from above downwards, and is not separated from that for the cuboid. The navicular has been lost in this tarsus. The shape of the ectocuneiform closely resembles that of J7esonyx ; it it is divided proximally by two facets, which form a right angle to each other ; the smaller and external is for the cuboid, and the 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 37 internal that for the navicular ; the posterior tuberosity bordering above the groove for the ‘peroneus longus’ is very large and much extended behind ; this is a character common to digitigrade forms like Fe/is, but absent in Ursus. ‘The mesocuneiform is high, slender and nearly as long as the ectocuneiform. The ento- cuneiform is elongated and broad, the posterior facet for the Mt. I is large and deeply concave ; this bonein Dyssacus is flatter and larger than in the Bear. Only the first metatarsal is pre- served, and it shows that the hallux was of good size in this type. Portions of vertebree, and especially of the caudals, were found with this skeleton (No. 777). The latter are large and much elongated, thus demonstrating that this form had a long tail. The two skeletons of Déssacus in the collection vary much as to the lengths of the same bones, but not more than in skeletons of recent Carnivores. Affinities of Dissacus.—This important discovery of the greater part of the skeleton of Drssacus adds much to our knowledge of the relationship of this genus to its probable successors in the Wah- satch and Bridger, Pachyena and Mesonyx. The superior molars of Dissacus are an exact counterpart, ona smaller scale, of those of Pachyena, although we observe that in D. carnifex the last supe- rior molar is more reduced than in the two known species of Pachyena. ‘The upper true molars of Pachyena still have the metacone smaller than the paracone, more especially marked in P. gigantea. ‘The inferior true molars of Pachyena are interme- diate in structure between those of D¢ssacus and those of JMWeso- myx; this is shown in the reduction of the metaconid; but the relative sizes of the other cusps, as compared with D¢ssacus, are the same. In JVZesonyx, on the other hand, the two external cones of the upper true molars are equal in size, and the last upper tooth of this series has been lost. The known species of Pachyena show no reduction of this tooth ; accordingly another species remains to be discovered in which this tooth is well reduced. As already remarked, the last upper molar of D. car- nifex is much smaller than the second, and this is what we should expect to find in an ancestor of JZesonyx. The presence of the metaconid in a rudimentary condition on inferior Mz, and some- times on M3 in Dissacus, proves that the Dissacus type of lower 38 = Bulletin American Museum of Natural History. (Vol. V1, molar has been derived from a typical tuberculo-sectorial tooth.’ As the AZesonyx type of lower molar is probably a degeneration from the less specialized tooth of Déssacus, so we must concede this to be derived from a tooth with a well-developed trigonid. Such an ancestral type of molar is found in Sarcothraustes, where all the cusps of the trigonid are nearly all equally well developed, but already in this genus the superior molars are completely tritu- bercular, with both external cones equal in size, as in AZesonyx. A comparison of the structure of the manus and pes in Déssa- cus with that of AZesonyx shows how closely these two genera are related. The position of the centrale in J/Zesonyx is quite differ- ent from that of Ayenodon, and resembles that of Dessacus. All the carpal elements in ussacus very closely resemble those of Mesonyx. It is quite remarkable to find in such an early type as Dissacus that the manus has undergone a considerable degree of ‘ displacement,’ as shown in the alternating articulations between podium and metapodium, indicating that Dzssacus led up to a digitgrade type ; in which there was a reduction in the number of the toes, as is observed in JZesonyx. ‘The structure of the pes and the relative lengths of the bones of the hind limb to each other, demonstrate that D7ssacus was a semiplantigrade form ; never- theless the calcaneum is much compressed and lengthened, and indicates the direction in which the foot structure of Déssacus was tending. The flat trochlear surface of the astragalus and the large astragalar foramen are typical of the plantigrades of the Puerco; again, as in Pachyena and Mesonyx, Dissacus has the large astragalo-cuboid articulation. Summing up the changes through which the D¢ssacus-Pachyena-Mesonyx line bas passed, we emphasize the following : 1. Growth of metacone of superior molars, and reduction of the last upper molar. tN Degeneration of the metaconid of the lower true molars, which is found well developed in D¢zssacus, and reduction of paraconid. 3. Change from the semiplantigrade condition of Déssacus to the digitgrade of A7esonyx. 1{See Scott, Uinta Mammalia, Trans. Am. Phil. Soc., 1889, p. 473. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 39 Note as cases of persistence the very close resemblance in structure of the carpus in Dyssacus and AZesonyx, the displace- ment of the metacarpus upon the carpus in D7ssacus, and also the articulation between the astragalus and cuboid in the latter. An undiscovered species of Pachyena, closely related to the P. ossifraga, but with the last upper molar more reduced than in that species, formed the transition stage between Déssacus and Mesonyx. Family PROVIVERRIDE Schlosser. Leptictide COPE, in part. Genus Deltatherium Cofc. Dentition: I$, Ct, P3,.M3. Superior molars with external cusps removed inwards from the external cingulum. No intermediate tubercles. Protocone large and V-shaped. A postero-external trenchant surface extending from the metacone. Last inferior premolar nearly molariform in structure ; true molars with trigonid high and trenchant. Inferior diastema large. Fig. 10. Deltatherium fundaminis. Right lower jaw, external and internal view. Natural size. 40 Bulletin American Museum of Natural History. [ Vol. VII, Deltatherium fundaminis Cofc. This is one of the most abundant types in the Puerco, and is represented in the collection by a number of examples, the best preserved being Nos. 780, 781 and 783. It is very instructive in showing how specialized some of the Puerco Creodonts were. In fact, the high differentiation of the carnivorous mammals of this forma- tion is surprising. D. fundamintis has already lost the first premolar in both jaws, and anterior to the second in Fig. 11. Deltathertum funda- c 5 : minis. Superior molars, crown view. the lower jaw there 1s along diastemes Netieiiee: The character of the true molars, and especially the trenchant form of the lower molars, is very different from that seen in most of the Creodonts of the Puerco. Cope’s material of Delfatherium is so well preserved that we are unable to add anything to his full description of this species. Deltatherium is closely related to the Wahsatch genus Szzopa (= Stypolophus), but is in some respects rather more specialized than that genus. 4. Order TILLODON Ex The relationships of the heterogeneous members of this order require careful consideration which we have not yet been able to give. Cope places Onychodectes and Conoryctes with the Creo- donta, but they show unmistakable affinities with Asthonyx and Tillotherium. Genus Onychodectes Cope. Onychodectes tissonensis Co/e. The collection contains a well-preserved skull and lower jaw (No. 785) of this species in which the teeth are badly worn. Another specimen (No. 786) consists of a part of the lower jaw containing the roots of all the premolars and the first two true molars. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 41 This skull is of great importance, as it is the most complete one ever found of this type in the Puerco. ‘The teeth agree pre- cisely in size with the type upper molars of O. tssonensis Cope. The skull is about as large as that of a small Didelphys. It is much lengthened between the glenoid facet and the last molar. The cranium is long and narrow, and there is no depression be- tween the cranial and facial portions. There is a very faintly developed sagittal crest, which extends as far forwards as the Fig. 12. Onychodectes tisson2msis. Skull and lower jaw, side view. Natural size. posterior boundary of the orbit. The nasals are narrow and elongate, and the anterior nares are terminal in position. The palate is long and narrow, and the palatines and pterygoids form very narrow posterior nares quite different from that of the Lemuroidea. The upper teeth are mostly broken off. The fangs of the ante- rior teeth indicate that there is a well-developed incisor shortly in front of the canine; the latter tooth is laterally compressed, and the first premolar is small and single-rooted. The second and third premolars are double-rooted ; the fourth premolar is three-rooted. It is evidently nearlyas large as the molars. There is no preglenoid ridge. The angular region of the lower jaw is partly preserved, showing that the condyle is obliquely trans- verse ; the coronoid is rather broad and the posterior border of the angle extends backwards. The inferior premolars are not 42 Bulletin American Museum of Natural History. |Vol. VU, spaced, and the posterior members of this series are robust. These have, however, been described and figured by Professor Cope. Onychodectes rarus, sp. nov. A prominent external cusp on each lower true molar, placed between the outer lobes. This new species is established upon a jaw fragment which con- tains two of the lower true molars (No. 824). The most striking character is the very prominent cusp which is placed upon the external side just in front of the posterior lobe. The trigonid is well raised above the talonid. The paraconid ite B is well developed; the protoconid is rela- tively robust and placed at the apex of the triangle and at an equal distance between the para- and metaconids. The talonid is broad and deep and extends into a basin on : : the inner side. The external interlobular Fig. 13. Onychodectes cusp of the second molar is smaller than that varus. Fragment of lower jaw with two true molars, of the first ; it arises from the base of the external view. Natural size. hypoconid, and is placed just opposite the convexity of the latter. Measurements. Psittacotherium multifragum Co/e. The division of the Tillodonta to which this species belongs is represented by a number of specimens, the best example of which is a nearly complete lower jaw (No. 754) associated with frag- ments of the skull, and with a number of teeth. Two much- worn premolar teeth are in place. As Cope has shown, the homologies of the lower cutting teeth are doubtful. The alveoli correspond with the formula given by Cope, I5, Cl, P3, M3. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 43 The portions of the skull preserved correspond somewhat with those figured by Cope of Hemiganus. They represent the top of the cranium and the upper and anterior border of the orbit. The cranium is compressed above, but like Oxychodectes, has no dis- tinct crest; the indications are that it was very long and narrow with an extremely small brain ; anteriorly the lambdoidal crests diverge very gradually instead of sharply, as in Hemiganus. They are very heavy and obtuse. There is no post-orbital process, and quite close in front of the orbit we observe as an exceptional feature a double infraorbital foramen. Other specimens related to these types are Nos. 755, 756, 757, consisting mainly of fragments of teeth and of bones. 5. Order AMBLYPODA Cope. This order of Ungulates includes the three suborders: Tali- grada (Cope). of the Puerco; Coryphodonta (Marsh) of the Wahsatch ; Dinocerata (Marsh) of the Bridger. Suborder TALIGRADA Cope. Primitive Amblypoda. Superior molars triangular, with selenoid cusps. Plantigrade. Astragalus with a distinct neck supporting navicular facet. A tibiale. Family PANTOLAMBDID.® Cope. Genus Pantolambda Cope. Dentition: I$, Ct, P+, M3. First upper premolar one-rooted ; second, third and fourth three-rooted, with internal cones. Canines laterally com- pressed. Pantolambda bathmodon Coe. No diastema in the dental series. These very primitive members of the Amblypoda are distin- guished by the following characters, as observed in an unusually perfect skull (No. 964) in this collection. The dental formula 44 Bulletin American Museum of Natural History. |Vol. VU, is typical ; the peculiar features of the superior molars are that although they present a broad transverse triangle, the apices of the three primary cusps (protocone, paracone and metacone) are brought close together as in the Periptychus, while the outer wall is very broad, exhibiting a parastyle and a metastyle, both well developed, while the mesostyle is feeble ; the intermediate conules are also feebly developed or absent. ‘The third superior molar exhibits a very large parastyle, making the outer border asymmetrical and foreshadowing the oblique development of the outer wall of this tooth in Coryphodon. ‘The first upper premolar is single- rooted, while the second, third and fourth each have Fig. 14. Pantolambda bathmodon. Crown view of superior molars. Natural size. three roots 5 and, although the crowns are wanting, this demonstrates the presence of a strong internal cone. The fourth premolar exhibits a single deeply crescentic external cusp (protocone) and a strong crescentic internal cone (deuterocone) with feebly marked conules. The canines are directed outwards and laterally compressed. The dental series is continuous, as in the type of this species, while in the larger species, ?. cavirictus, there is a considerable diastema behind the canines. The skull is of a very ancient type, exhibiting the following primitive characters: The anterior nares are terminal in position ; the front border of the maxilla descends vertically, and the pre- maxilla, which is broken away in these specimens, was apparently short. ‘The cranium is twice as long as the face; the brain-case proper is low and broad transversely; it is surmounted by a sharp sagittal crest and flanked posteriorly by lateral occipital crests; the occiput is, therefore, very broad and low, as in Perip- tychus, in lateral view. We observe that the zygomatic arches are very slender, and there is a wide space between the postglenoid process and the posttympanic. The posttympanic and paramas- toid processes are confluent and very sessile. ‘The basal view of 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 45 Fig. 15. Pantolambda bathmodon. Skull: lateral, dorsal and ventral view. One-third natural size. 46 Bulletin American Museum of Natural History. (Vol. VII, the skull shows that the pterygoids are extended very far back. The postglenoid processes are very small; in fact, this view brings out well the simple and undifferentiated character of the base of the skull. The posterior border of the lower jaw descends vertically behind the condyle, as seen in specimen No. 962, which probably belongs to this species. The scapula (No. 964) exhibits a shallow glenoid cavity, close above which is the base of the spine; the neck, there- Se ee fore, is extremely short ; there is a Sa long coracoid process recurved dis- tally. The humerus is massive ; it is char- acterized by a very large and promi- nent deltoid crest, which extends below the middle of the shaft ; on the inner surface of the shaft is a slightly prominent crest for the flexor muscles ; there is a large entepicon- dyle perforated by a foramen, and upon the outer side of the distal extremity there is an acute ridge which we observe is not developed in Periptychus ; distally the humeral condyies do not display any intertro- chlear ridge. The ulna is placed en- tirely behind the radius ; its proximal section is deep anteriorly ; posteriorly jie pate ee the radius is preserved, but is so pete Rurnetis: patency Cr much damaged that its characters cannot be made out. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 47 6. Order CONDYLARTHRA Cope. Dentition bunodont. Manus and pes pentadactyl. Elements of carpus and tarsus serially arranged. Humerus with an entepicondylar foramen. Femur with a third trochanter. The suborder Condylarthra was established by Cope’ to in- clude the genus Phenacodus. At that time he considered this genus to bea Perissodactyle, and placed the Perissodactyle group as an order, including the suborders Diplarthra and Condylar- thra. Later® he proposed the order Taxeopoda, to include the Proboscidea and Condylarthra, but in his paper ‘ On the Classifi- cation of the Ungulate Mammalia,’* removed the Proboscidea from the Taxeopoda and gave it an ordinal position; in the same paper Cope included under the Taxeopoda the suborders Condylarthra and Hyracoidea. DoUBTFUL POSITION OF THE PERIPTYCHIDA. In the present state of our knowledge it is difficult to say what forms should be included in the Condylarthra. If we adhere strictly to the diagnosis of this group laid down by Cope, we should have to omit the Periptychidz front this suborder, because in the genus Leriptychus the tarsus ts not serial; there is a displacement of the astragalus upon the cuboid, and the whole structure and angulation of the hind foot is different from that of the type genus Phenacodus. Pertptychus is quite as closely related in its pes to the Amblypoda as to the Condylarthra. Periptychus has the simple bunodont dentition of the Condylarthra, but it has the strictly trigonal molar of the Amblypoda. The most specialized family of Cope’s Condylarthra is the Meniscotheriide. Osborn* has shown that this is analogous to Chalicotherium in Cope’s Ancylopoda. It thus appears possible that the Periptychide and Meniscotheriidze must ultimately be 1 Am. Nat., 1881, p. ror8. 2 Am. Nat., June, 1882. 3 Proc. Am. Phil. Soc., 1882, p. 438. 4 Am. Nat., 1892, p. 507. 48 Bulletin American Museum of Natural History. [ Vol. VII, removed from the Condylarthra, and that the Condylarthra may ultimately include only the stem forms of the Artiodactyla and the Perissodactyla. At present we enlarge the order by adding to it certain forms which Cope has placed among the Creodonta; we thus transfer the genus A/voclenus and family Mioclenide. We agree with Schlosser and Scoit that the structure of the teeth in this genus shows it to be more closely related to primitive Ungulates than to any of the Creodonts. . The following table will illustrate the arrangement and sub- divisions of the Condylarthra proposed in this paper. Family MIOCLEANID/S, fam. nov. Genus Mioclzenus Coe. Dentition : 71, C1, Pi, M3. Third and fourth superior premolars with single internal cones. Superior true molars tritubercular, with hypocone very rudi- mentary. Last upper and lower molars reduced. Inferior premolars much enlarged and very simple in structure. Inferior true molars without paraconid. The genus AZioclenus was established by Cope,’ the type species being 17. ¢urgidus. At the time of the description of this genus Cope considered it closely related to Luprotogonta (=Pro- togonia) and in his divisions’ of the Condylarthra in 1881 placed Mioclenus in the family Phenacodontide. Later® he omitted this genus from the latter family, saying: “ I believe it to be Artiodac- tyle.”’ Upon the discovery of the structure of the skeleton of J7Zzo- clenus ferox Cope associated this species with 47. turgidus, and referred both species to the Creodonta. ‘The JZ. ferox has since been raised to generic rank by Scott as the type of C/enodon. Cope in his ‘ Tertiary Vertebrata,’ and later in his ‘ Synopsis of the Puerco Series,’ included a great many other species under the genus Mioclenus, but Scott’ in his paper ‘A Revision of the 1 Proc. Hee Phil. Soon iSepu 17, Tea ae 2 Am. Nat., 1881, p. 1018. 8 Proc. Am. Phil. Soc., Dec. 16, 1881. 4 Proc. Acad. Nat. Sci., Phil., 1892, p. 321- 49 Osborn and Earle, Fossil Mammals of the Puerco. . wW ima) ioe) cl ‘apeisiqueyd-rwoes saq ‘JaoRj [eauLO]eO-o[NGY yW ‘payesojiad snyeseijsy ‘e[Ajoeposstiag Ul Sv SIP] -oweid jo uoljeUojsuely ‘payso10 SIzjow JaMOT ‘sapjoqn} jeusoUI pue sjusosaid Jeuta}xXo YIM siejow addy, ‘peoiq opsuery sang ‘wpusayjossiuapyy “Vv ‘ApRASISIP sod puv snuvyy “joory [eau Teo-ornqy ON ‘paaoois snjeseijsy ‘sniswny jo a0vy iolsajsod uodn AyTuo Uy sivjoulsid jo uoljewIojsuerd y, ‘ajeipenb sivjow isddq ‘peoiq 9o[sueij AWG ‘aye10j1od ut ‘e[Ajoepossiiag Ul Se ‘wpyuopospuayg “© “sIv[OU 1aMO] JUOpOUaTas-oydo] puv ‘raddn yuopouajas-ounq yuA\\—'g ‘poqyeioyiod ‘yey | snyeseysy ‘vayyoo.s) jerowny jo 90ey Jolejue uodn vu ‘sjaay = yyIM «sivpowaid = s011 -aJuy + ‘“pasiepua sivjow J9MOT pue soddn yjimoj pue pry ‘(auoooddy =pue a] A}soj0.1d) sajnsuio yeuseyur Areyuoweyd -dng ‘ajeipenb saaou ‘Arjour -wAs Jepnsuel YM siejowW ‘srvjou ioredns ut passaid -WOOd 9[suRLy = saATTIWIIG ‘apy mggidag “% ‘UMOUYUN UO}e -[aYS ‘spaey Suryory srepoussd yeusojut = Azeyuowayd ‘ueyd ! sdsno -dns Suryory siepoyy ul wprympgidag 94} OF IR] -1tuIs sivjowaid pure siejoy ‘mpunjIoiy “1 “sazjow Jamo; pue saddn yuopoung yitA\—'* “VYHLAVTIAGNO,) [|Warch, 1895.) 50 Bulletin American Museum of Natural History. |Vol. VII, Creodonta,’ has removed many of the species to a distinct generic position. He says of M/voclenus: “ The name AMtoclenus should be restricted to those forms which agree with the type species JZ. turgidus in the extremely broad, low and massive premolars, which equal or exceed the molars in size,” etc.; and later remarks, “Tf, as Schlosser has suggested, it becomes necessary to refer Mioclenus to that group [Condylarthra], it will form a very dis- tinct family of that order.” Mioclzenus turgidus Cope. First superior true molar with a rudiment of a hypocone. Last superior and inferior true molars reduced in size. Inferior true molars without a postero- internal cone (entoconid). First inferior premolar spaced. There are numerous fragmentary specimens (Nos. 921-936, 938, 939) of this species in the collection, the best preserved being Nos. 930, 921, 922, 933. These specimens together illustrate the structure of the greater part of the dental series. Associated with No. 921 are fragments of the skeleton, especially a well-preserved sacrum. G Deh Stk oe, bo Oe _ Fig. 17. Mioclenus turgidus. Superior and inferior molars. Crown view. Natural size. (No. g21.) Dentition—The second superior premolar consists of a single cone, without heels. The third and fourth have well-developed internal cones, which are single. These teeth have no interme- diate tubercles. In the specimen under description (No. 921) there is a large diastema in front of the first superior premolar, but whether this interval is natural or not remains to be deter- mined from better material. The superior true molars are very primitive in their characters, more so than in any of the known 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 51 Condylarthra ; the first molar presents rudiments of the parastyle and mesostyle on the outer wall; the molars have well-developed intermediate tubercles, and only on the first is there a trace of ahypocone. The third true molar is much reduced in size, and the second is larger than the first. The reduction of M3 is against the Condylarth affinities. The shape of the inferior premolars is highly characteristic; the first is not preserved, although an impression of its crown is left on the specimen ; it was simple in structure and isolated from the small canine in front, and also from the second premolar behind. The second, third and fourth premolars have much enlarged and swollen crowns, with only slight indications of pos- terior heels and with no trace of a deuteroconid; the third and fourth present rudimentary anterior and posterior basal cusps. The inferior true molars closely resemble those of Euprotogonia ;' the crowns are very low, broad ; and the trigonid is somewhat raised above the talonid. In Cope’s description of the dentition of MW. turgidus he describes a trace of a paraconid on the first and second molars. In the American Museum specimens of this species the second true molar has a trace of a paraconid, and this tooth, as well as the third molar, exhibits no entoconid. The sacrum is broad and short, its antero-posterior and trans- verse diameters being about equal. In its general characters it closely resembles that of Oreodon. The neural spine and prezyga- pophyses are low. In contrast with the sacrum of the Carnivora we notice the position of the sacro-iliac attachment; it is elon- gated, narrow and parallel with the antero-posterior axis of the sacrum. The surface for articulation with the ilium is confined to the transverse process of the first sacral vertebra, as in the Ungulata in general. SYSTEMATIC POSITION OF MIOCLANUS. The most striking character of the dentition of J/zoclenus tur- gidus which points to its relationship to the Periptychidz is the enlargement of the lower premolars. The absolutely tritubercular superior molars, without a hypocone, except on Mz, prove this 1 Am. Nat., April, 1893. 52 Bulletin American Museum of Natural History. (Vol. VU, form to be the simplest and most primitive type in its tooth structure of any of the known Condylarthra. A specialization in this genus—a character not anticipated in so old a type—is the probable presence of a diastema in the dentition ; this is an unusual character for any Puerco Ungulate to exhibit. One of the Creodonta (De/tatherium) of this formation is also quite special- ized in this respect. The discovery of fragments of the skeleton of A/zoclenus tur- gidus is of great importance, and a well-preserved sacrum, already described, adds much weight to the theory of the ungulate affinity of this genus. M. Pavlow' has suggested that J/. zurgédus is an intermediate form between Periptychus rhabdodon and Antsonchus sectorius. It appears rather that J/ioclenus is much more primitive in its dental characters than Pe7/ptychus, and should be placed below that genus structurally. As JZ. ¢urgidus is rather an unspecial- ized type in its dentition, it is possible that it may have been one of the few types of the Puerco which persisted in later periods, and Earle’ has suggested elsewhere that this genus may stand in ancestral relationship to some of the White River bunodont Artiodactyles, such as Leptocherus. Family PERIPTYCHIDE Cope. It is convenient to divide this into two subfamilies : 1. Amzson- chine, to include the smaller and more primitive forms; 2. Periptychine, to include the larger and more specialized forms. 1. Anisonchine. 2. Periptychine. Smaller forms. Superior molars Larger forms. Superior molars with with intermediate tubercles (conules) | conules well developed. Interior mo- suppressed or wanting. Inferior mo- | lars with paraconid well developed. lars with paraconid reduced or want- | Astragalus with a short neck. ing. ? Astragalus with elongate neck. 2 Etudes sur Histoire palaontologique des Ungulés, Bull. de la Société Imp. des Natu- ralistes de Moscou, 1887. p. 19. 2 Science, July 28, 1893, p. 51. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 53 Subfamily PERIPTYCHIN. Genus Periptychus Cope. Dentition: I, C+, P#, M3. Teeth vertically sculptured. Protocone of superior premolars much elongated and recurved. Last three superior premolars with crescentoid internal lobes. Superior true molars with supplementary in- ternal cusps well developed. Inferior true molars with paraconid. Periptychus rhabdodon Cofe. Superior true molars as broad as long, and provided with two intermediate tubercles ; third not reduced in size. Superior premolars much enlarged, and with internal cusps uniting into a continuous internal crescent. All the teeth strongly sculptured. A large number of specimens (Nos. 854-878) represent this species in the collection. The dentition of P. rhabdodon has been fully described by Cope. ‘ The best example of part of a skeleton in the collection is a hind limb with a well-preserved calcaneum and astragalus (No. 837). The femur is short and rather stout, the third trochanter is placed slightly above the middle of the shaft. The crest of the #éza is very prominent and extends far down upon the shaft ; the distal articular end of the tibia faces obliquely outwards, and is nearly plane ; there is a slight ridge dividing the internal from the external trochlea; the internal malleolus is very prominent and peculiar in form; it is strongly grooved for a flexor tendon. The fibula is well preserved ; it is a short and heavy bone. The proximal extremity is flattened and expanded, and it exhibits a concave facet for articulation with the tibia; externally this end has a prominent rugose process; the shaft at its middle part is oval in section, and its anterior face is separated longitudinally by a ridge ; the distal extremity is much enlarged and presents a very plane (articular) surface for the astragalus; the external mal- leolar tuberosity is strongly marked, and is nearly as prominent asin the Bear. As compared with that of Ursus, the fibula of P. rhabdodon, in contrast with the size of the tibia, is much larger, and its shaft is thicker. 54 Bulletin American Museum of Natural History. \Vol. VU, There are two astragali of P. rhabdodon in the collection, and in both there is a plainly marked astragalar foramen ; this aperture is situated well toward the median trochlear surface, and com- mences just at the posterior limit of the articular face. The pres- ence of this foramen in Periptychus is a constant character, and in this respect it differs much from the genus Coryphodon, in which it is variable; in both these genera the foramen has the same position, namely, between the ectal and sustentacular facets. We doubt whether it transmitted a flexor tendon, as it is not clear how a tendon could traverse this foramen and then pass out- wards under the sustentaculum ; it is more likely that this foramen transmitted a blood vessel or a nerve. We are not aware that it exists in any recent Ungulate, yet it is a constant character of all Puerco forms, and a vestige of it has.been observed by Wortman in the pinniped Carnivora. Measurements of bones of Hind Limb. Wenet hy often teeter lel-teletie leit liek 163 Width of same proximally........-......-- .046 Wength of tibiake- on Ria apete ist lemme rere ete .140 Breadthdistalllly,pyertetteictetetr-tteteletaiolelstel-tet rt .027 ene th rote tla vaperterte tokio oe ee Breadthidistallypne ener trl ten eter .O16 Total length of limb, allowing for ankle flexure, .280 Periptychus coarctatus Cofe. Internal cingulum of inferior premolars discontinuous. Superior premolars with great transverse extent. Intermediate tubercles present on true molars. Superior and inferior true molars with external cingulum. ‘The P. coarctatus is represented in the American Museum col- lection by the greater part of the upper and lower dentition of one individual (No. 850). This species is a decidedly smaller type than P. rhabdodon. The upper true molars are nearly the size of those of FP. brabensis, but their transverse diameter 1s greater ; the inferior premolars on the other hand, are relatively enlarged in P. coarctatus. Dentition.—The last superior premolar has a greater transverse extent than the first true molar; yet the superior true molars are also much extended transversely, their external cones are small 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 55 and considerably raised above the surface of the teeth. The in- termediate tubercles are well developed in this species, and the last true molar as compared with the first and second is relatively much smaller than in P. rhabdodon. ‘The inferior premolars are smaller than those of P. rhabdodon, in which the anterior and posterior tubercles are weakly developed. The first upper true molar is larger than the others, and its protoconule is more robust than in the allied species. The last lower molar is small, and the three cusps of the talon are more distinct than in P?. rhabdodon. Measurements of teeth of P. coarctatus. Length of last four upper molars..........-. .032 Length of superior true molars............. .022 Eengthiofintertor premolars... 22-62... 5. .'-:- .040 Mength of inferior true molars, ....-..:--.-- .028 Periptychus brabensis Co/c. Cingula of inferior premolars discontinuous. Transverse diameter of superior premolars less than that of true molars. Superior and inferior true molars with external cingula reduced or wanting ; intermediate tubercles of superior true molars wanting. There is only one specimen of this species in the American Museum collection (No. 849). This contains the greater part of the lower dentition and some of the upper molars. The external face of the molars in both jaws is only slightly sculptured. The first superior molar is triangular in outline, with protocone smaller than in the second. The second superior molar is well preserved, and is of a square form ; as compared with other species this tooth is considerably modified ; the protocone, instead of being a simple tubercle, as in the P. dradensis, is a crescent, and the intermediate tubercles are fused with its anterior and posterior spurs; both the internal supplementary cusps are rela- tively more developed than in the large P. rhabdodon. The last superior molar is not reduced in size as in the ?. coarctatus. The transverse diameters of the last upper premolar and first true molar are about equal. The last two tzferior premolars are much elongated antero-pos- terlorly, with a small transverse diameter; the anterior and pos- terior heels are prominent but not continuous internally. The 56 Bulletin American Museum of Natural History. \|Vol. VU, second inferior molar is smaller than the first, and in this respect the P. drabensis differs from the P. rhabdodon ; the paraconids of the lower molars are not so distinct as in the last-named species. A single zcisor is preserved with this specimen, which probably belongs to the lower series ; the crown is strongly compressed, with a slightly ex/arged posterior heel. Genus Ectoconus Cofe. Dentition: 1%, C+, P{, M3. Last three superior premolars with internal crescents ; third and fourth sub-molariform. Superior molars consisting of seven cusps and two external cingular cusps. Last inferior premolar with cusps of trigonid well developed. Inferior molars sextubercular, with an antero-internal accessory cusp ; internal to paraconid. Ectoconus ditrigonus Cope. Superior and inferior true molars with a strong external cingulum. Last superior molar nearly as large as first. Postero-external cingular cusp of supe - rior true molars opposite metacone. This species was first referred by Prof. Cope’ to Conoryctes, but was later’ established as the type of “Zcfoconus. ‘The dentition (Nos. 880-888) has been only partially described by Cope, and our abundant material enables us to complete his account. The superior incisors are of a simple conical form, and increase in size from within outwards; their position is peculiar, as they are sep- arated by an interval from each other. ‘The finest specimen of Ectoconus in the collection is No. 880; in this both upper and lower teeth are from the same individuual ; the upper canine was large, as is shown by the basal part of the crown, which is pre- served ; the single alveolus for the first superior premolar is intact, and this tooth probably had a simple crown; it is slightly sep- arated from the canine and the premolar succeeding it. The premolars of £. ditrigonus differ very much in their pat- tern from those of Periptvchus. ‘The last three superior premolars have one external cone, and this is not elongated or sculptured ; the internal lobes of these teeth are crescentoid, and in the last / 1 Am. Nat., 1883, p. 968. 2 Am. Nat., 1884, p. 796. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. Sy two of the series they early unite by wear with the two interme- diate tubercles, thus presenting a crescentoid tract of worn enamel. The last upper premolar is a much smaller tooth than the first true molar. A marked characteristic of this genus is that the third and fourth premolars remotely repeat the molar pattern, whereas in Periptychus they are wholly dissimilar. The molars are rectangular in form and are much drawn out transversely ; they therefore differ in shape decidedly from those of theallied genus Perzptychus. The large number of cusps is the most distinctive character. In strong contrast to other Puerco Ungulates, the upper molars of Lctoconus have an external cin- gular cusp placed just outside of the metacone, and whether it is homologous with the median (mesostyle) or the posterior cingular cusp (metastyle) of higher forms, it is difficult to say. In some examples of this species there are two cingular cusps on the second superior molar, one behind the other. The protocone of the molars is large and soon unites by wear with the intermediate tubercles ; these conules are larger than in Peviptychus, especially the protoconule. The two internal supplementary cusps (pro- tostyle and hypocone) are well developed, and arise from larger cingula than in Periptychus. The last superior molar of Z. dtri- gonus is nearly as large as the second, and is provided with two well-developed external cones; the paracone is connected with the external cingulum by an oblique ridge. In all the superior molars of this species the antero-external part of the basal cingu- lum is prominent, and on the last upper molar forms a very dis- tinct cusp. The inferior molars are much crowded together, and there is no diastema between the canine and the first premolar (No. 880) ; this tooth is absent in the specimen under description, but repre- sented by a small alveolus placed close to the second premolar. In Cope’s specimen of &. dtrigonus the second inferior premolar has a small internal cusp; on the third premolar this cusp is also present, and the two cusps of the talon are fully developed. The last inferior premolar (see No. 890) is quite complex in structure for so early a type; it exhibits the three principal cusps of the trigonid, but the postero-internal cusp is absent. The complex sub-molariform structure of the posterior and lower premolars in 58 Bulletin American Museum of Natural History. \Vol. VX], Ectoconus is to be noted in contrast with Periptychus. In No. 890 the fourth premolar is very similar to the first molar. The inferior true molars are broad, with a low and indifferent arrangement of the cusps; the protoconid is on a line with the paraconid, and not piaced opposite the interval between the two internal cusps as in Periptychus. Vhe specimen, No. 880, in the collection exhibits a well-marked cusp on the second and third inferior true molars just within the paraconid ; this cusp is not present in Periptychus. The last lower molar is considerably extended antero-posteriorly, and the hypoconid is placed far forward like that of Perzptychus. The hypoconulid is very large’ on this tooth, much larger even than the entoconid. Measurements of Teeth of Ectoconus. M. Total length of superior molars ..........-. .062 Length of superior premolars.........- soe KOR Length of superior true molars... ......... .029 Total length of inferior molars... .-...-....- .064 Length of inferior premolars: -.--.-=-...-.- .030 enethiof antenormolatsee pyri etl .034 Humerus.—Near the upper and lower jaws (No. 888) was found a large humerus from the left side, which may belong to this species, although the association is somewhat doubtful. It measures slightly over 150 mm. in length. It is robust, with much more prominent crests for the deltoid and supinator muscles than are seen in Periptychus. The latter or ectepicondylar ridge is exceptionally prominent and is carried over upon the front face of the shaft as in fossorial animals, and differing from both Periptychus and Pantolambda. Subfamily ANISONCHIN#. Genus Haploconus Cope. Dentition: 11, P#, M%: Fourth superior premolar only with an internal cone, Superior molars with protocone crescentoid in form ; no distinct inter- mediate tubercles ; no anterior supplementary cusps (protostyle). Last three inferior premolars with heels ; no internal cusps. Inferior molars without paraconid. 1895.]| Osborn and Earle, Fossil Mammals of the Puerco. 59 The genus Haploconus is readily distinguished from the three others of this subfamily: (1) By the absence of a deuterocone on the third superior premolars ; (2) the development of the hypo- cone is less advanced than in the genus An/sonchus, but more so than in Hemith/eus; (3) the inferior premolars of Haploconus are elongated, whereas in Hemithleus (H. apiculatus) they are con- siderably enlarged, more resembling those of Mvoclenus. Haploconus lineatus Cofc. Fourth superior premolar enlarged ; fourth superior premolar spaced with internal cone crescentoid. Hypocone of superior molars larger than anterior supplementary cusp (protostyle). Third and fourth inferior premolars some- what elongated and trenchant. The characters of the lower premolars in this species are closely related to those of Amzsonchus mandibularis, although Cope states that in the last-named species the third superior premolar has an internal lobe, which places it in Anzsonchus. The best specimen (No. 891) in the collection of AH. /incatus is a set of superior molars, associated with the fragmentary remains of a part of askeleton. The structure of the skeleton in Haflo- conus has been until the present time totally unknown, and we are happy to be able to give some: information as to it. Many broken fragments of the long bones are associated ; they are long and slender, and their general proportions are as in the limbs of Lemur varius. A distal extremity of a humerus exhibits a well- marked entepicondylar foramen. A proximal part of an ulna exhibits an olecranon process which is long, slender and strongly compressed ; the coronoid process is short and does not extend as far forward as in Ungulates ; the two divisions of the sigmoid cavity are unequal in size as in the Carnivora, and lastly, the ulno-radial facet is small and limited to one side as in the Carni- vora. The most interesting bones are, however, part of a calca- neum and an astragalus. The proximal portion of the calcaneum is high and narrow, the ectal facet is raised high above the sus- tentaculum ; this latter facet is round and oblique in position. The astragalus differs widely in its characters from that of Perip- tychus ; the trochlear surface is nearly plane, there being only a 60 Bulletin American Museum of Natural History. (Vol. V1, slight median depression ; a large astragalar foramen perforates this bone at the posterior limit of the articular surface ; unfor- tunately the navicular facet is missing, but enough remains of the base of the neck to show that the latter was slender and elongated, in marked contrast with the short neck in Periptychus. We conclude from these characters of the skeleton that Hap/o- conus lineatus was an exceedingly slender type, with elongated limbs, probably adapted for an arboreal life. These characters are also radically different from those of the heavy-limbed plan- tigrade Perzptychus, which has hitherto been supposed to be closely related to Hafloconus when judged by the teeth alone. Genus Hemithlzus Cofc. Dentition : C1, P{, M3. Third and fourth superior premolars with single internal cones; protocones of superior molars crescentoid ; supplementary cusps (protostyle and hypocone) of equal size. Inferior premolars with enlarged protoconids; deuteroconid variable. Paraconid reduced in molars. Hemithlzus kowalevskianus Co/c. Anterior and posterior cingula of upper molars slightly produced into sup- plementary internal cusps ; third upper molar much reduced in size. Third and fourth inferior premolars conic and not larger antero-posteriorly than true molars, and without anterior tubercles; heels small. This species is represented in the col- lection by a fine set of upper and lower true molars, with part of the premolar series (No. goo). This species is smaller than /. apiculatus, and its inferior pre- molars are simpler in structure and slightly Fig. 18. Hemithleus kow- 2 : aleoskianus. Crown view of enlarged, although less so than in A/zoclenus superior and inferior molars. = Natural size. turgidus. Genus Anisonchus Cofe. Last two superior premolars with internal cones. Superior molars quadrate in form, witha large development of the hypocone. Inferior premolars much elongated antero-posteriorly, with the third longer than the last. Paraconids of inferior true molars present but reduced, 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 61 Anisonchus is the most highly developed member of the sub- family, and the superior molars of this genus are transitional in structure between the tritubercular and quadritubercular forms. The hypocone is largely developed and extends inwards beyond the protocone. The inner cones of the upper premolars are cres- centoid in section. The structure of the superior true molars in Anisonchus is more advanced than in the allied genus Hemithleus. We believe the latter generic name should be retained. Anisonchus mandibularis Cope. Third inferior premolar much elongated antero-posteriorly. Hypocone of superior molars more extended inwards than in A. sectordus. A number of mandibuli represent this species in the collection (Nos. 893, 894, 895 and 896). It is uncertain whether the 4. man- dibularis is fully distinct from A. sectorzus, as in both species the third inferior premolar is elongated. A mandible in the collec- tion (No. 896) has the entire inferior dental series finely pre- served ; the canine is long and slender; it is convex externally and concave internally. The canine is separated by a diastema from the first premolar, which in turn is also some distance from the second. The first premolar is a very minute tooth, and is much smaller than the second. The following specimens belonging to members of the Anison- chine have not yet been determined, owing to their fragmentary characters : Nos. 893, 897, 903, 904, 907, 917. RELATIONSHIPS OF THE GENERA OF THE PERIPTYCHID. The question of the relationship between the different genera of the Periptychide is a difficult one to decide. In most cases we know very little about the skeleton of these forms, and there- fore must depend upon dental characters to establish their affinities. Mioclenide.—As we have already attempted to show, we con- sider Mioclenus turgidus more closely related to the Periptychidz than to any of the Creodonta. The characters of the teeth show it to be the most primitive member of the suborder. The 62 Bulletin American Museum of Natural History. [Vol. VU, superior true molars in J. ¢urgidus are simpler.in structure than in most of the genera of the Periptychide ; they are tritubercular without any internal supplementary cusps ; the external cones are low and conical, and the intermediate tubercles are well developed. The last two superior premolars have one internal and one external cone: by the presence of an internal cone to superior Pm. 3, M. turgidus is more advanced in this single character than Hap- /oconus, otherwise we must consider it to be a more primitive form than the latter genus. The lower premolars in JZ. turgidus are very simple in structure and much less complicated than those of Haploconus, approaching in their structure those of Hemzthleus. The inferior true molars of JMJvoclenus also remind us of Luprotogonia. Antsonchine.—In this family there are ten or twelve distinct species, partly representing different lines of descent and partly different stages of evolution in the same line. It is evident that this evolution and divergence represents a long period of time, but unfortunately we have few data as to the vertical distribution of species. Cope has unfortunately made the successive addition of internal cones to the premolars the sole basis of his generic definitions—by analogy with other groups this merely indicates successive modifi- cation tn time of animals perhaps belonging to different phyla. The true key to the separation of the phyla is not this character chosen by Cope, but the rounded or flattened form of the upper and lower premolars. ‘Taking this key, we discover two sharply defined series, as follows: Series A.—In which the lower pre- molars are flattened and develop anterior basal cusps, while the upper premolars exhibit crescentic internal cones. Series B.—In which the lower premolars are rounded and never develop anterior cusps, while the upper premolars exhibit comzc internal cones. The natural inference is that these two series represent two dis- tinct or divergent lines of descent, and that the parallel successive stages of modification in time are indicated by the gradual addi- tion of secondary cusps upon the premolars and molars. A reclassification of Cope’s species on this basis would greatly sim- plify our conceptions, but would introduce endless confusion in the nomenclature. 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 63 A. [Related to Periptychus. | B. [Related to Zctoconus. | Premolars Lower premolars flat- | Lower premolars round- lu pper mo- to which cusps | tened. Internal cusps of | ed. Internal cusps of lars; num- are added. upper premolars cres- | upper premolars conic. |berofinter- ate centic. nal cusps. Upper. | Lower. P3-4 | P2-4 | Hemithlzeus apiculatus;. H. kowalevskianus. 2 H. corniculatus. = a Anisonchus gillianus. | I < P3-4 oe sectorius....| A. coniferus. | I cs S oo mandibularis. | I P4 P4 | Haploconus xiphodon...| H. cophater. H. ento- | I conus. ae Oo ae | lineatus. The vertical lines of species as here arranged do not imply phyletic descent, for it is noteworthy that the species of ap/o- conus (otherwise the most primitive) usually lack the paraconid in the lower molars, while the species of Azésonchus and Hemith- /eus (otherwise more specialized) usually exhibit the paraconid, which is always to be considered a primitive mark. The table does illustrate the parallel transformation of species in different phyla both in the addition of internal cones to the premolars and of internal styles to the molars. ‘The species .of Series A are evidently most nearly related to the still more highly Peripty- chine, for Hemtthleus apiculatus, with the addition of an inter- nal crescent to the second upper premolar, would closely resem- ble a miniature Periptychus. On the other hand, the species of Series B are related to Actoconus. Thus, Series A represents one line, Series B a second line, and Zefodon a third. RELATIONSHIP OF THE PERIPTYCHID# TO OTHER GROUPS. It is probable that none of the known genera of the Peripty- chide, with the possible exception of MJvoclenus turgidus, per- sisted into higher types. This is indicated in Perzptvchus by the peculiar specialization of its premolars, andif this genus is related to any Wahsatch group it is to the Amblypoda. We do notagree with Schlosser in deriving any of the Artiodactyla from the Anisonchine, which were already somewhat specialized in their structure, and were probably adapted for an arboreal life, repre- 64 Bulletin American Museum of Natural History. (Vol. VU, senting in the differentiation of the Puerco fauna the small and agile climbers. As Earle’ has shown elsewhere, it is probable that in the Puerco we have a stem form of the Artiodactyla in Proto- gonodon, a genus which was more closely related to Euwprotogonia than to any of the Periptychide. Family PHENACODONTID. Genus Euprotogonia Coe. Protogonia COPE. Last superior premolar with only one well-developed external cone, trito- cone rudimentary. Superior true molars sextubercular, and without parastyle or mesostyle. Last lower premolar with deuteroconid and paraconid; talon large. Inferior true molars generally quadritubercular, with hypoconulid. Cope’ has only recently substituted the name Lwufrofogonia for Protogonia. The type species is Z. subguadrata,;’ this differs considerably in the structure of its upper molars from £. puer- censts. Cope’ in his revision of the species of Huprotogonia has omitted the type species. In the type species the section of the external cusps of the upper true molars is lenticular, and the hypocone is very rudimentary. We therefore consider the type form nearer to A/zoc/enus than to Luprotogonia. The most common form of this genus from the Puerco is the LE. puercensis, and it illustrates the characters of the teeth in this genus. The tritocone, or second external cusp, on the fourth superior premolar is variable, but in &. puercensis this cusp is small and well defined. ‘The most important generic characters distinguishing Euprotogonia from Phenacodus are the conic form of the external and internal molar cusps, and the absence of the cingular cusps. The latter are always present in Phenacodus, but the species of this genus do not always exhibit both anterior and median cingular cusps. The last inferior premolars in Zupro- togonia and Phenacodus are similar in structure; in the former the metaconid is less separate from the paraconid than in the latter. 1 Am. Nat., April, 1893, p. 377- 2 Am. Nat., 1893, p. 378. 3 Proc. Am. Phil. Soc., 1881, p. 4 Synopsis of the Vertebrata ofthe Puerco Series, Proc. Am. Phil. Soc.. p. 359, 1888, 1895.] Osborn and Earle, Fossil Mammals of the Puerco. 65 In &. puercensis the inferior true molars are quadritubercular, and there is no paraconid. In £. p/ictfera this cusp is present, as described by Cope. Comparing the development of the inferior crescents in Luprotogonia with Phenacodus, we observe that in the former genus they are fully as well marked as in the latter. Euprotogonia puercensis Cope. Last superior premolar with a small tritocone. Superior true molars with hypocone smaller than protocone. Inferior true molars lacking the paraconid. This species is represented in the American Museum collection by a number of specimens (Nos. 940-947); the most perfectly preserved is No. 941, in which both upper and lower teeth are from the same individual. Dentition—As shown by Cope, the characters of the upper true molars of £. puercensis approach more closely those of Phena- codus wortmani than of P. primevus. It appears that if we con- fine ourselves to the premolars in separating Huprotogonia from Phenacodus, there are no distinct lines between the two ; the case with the true molars is different. In 7. priémevus the external lobes of the superior molars are distinctly flattened, and there is a well-marked para- style and mesostyle. The ex- ternal lobes of the upper molars of Phenacodus wortmant are conical and closely resem- ble those of &. puercensts. The last inferior premolar in the latter is as complex as in d4 1 2 Fig. 19. Euprotogonia puercensis. A,Crown Phenacodius, the two posterior view of superior molars. Ar, Last inferior pre- z molar, and # first two inferior true molars and cusps of the trigonid being last milk molar. A and A1, are from the same r individual. Natural size. equal in size, and the talon having the same structure. It is of importance phylogenetically to recognize the rather complex structure of this tooth in Eupro- togonia as a Perissodactyl ancestral type, because in Protogonodon, a supposed Artiodactyl ancestor, the last inferior premolar is simple. [Aay, 7895.] a) 66 Bulletin American Museum of Natural History. \|Vol. VII, AFFINITIES OF EUPROTOGONIA. First : it is remarkable that the grinding teeth of 2. puercensis agree precisely in size and in almost every detail of structure with those of Ayracotherium vulpiceps Owen,’ from the London Clay. The latter is only a shade more modernized. This strongly confirms Schlosser’s supposition that this species is a direct ancestor of the Equidz. Second: it is significant that this genus is the only one known from the Puerco which has a well-developed sextubercular superior molar. In some of the Periptychidee (Anisonchus) the hypocone is large, but the intermediate tubercles are absent. In Huprotogonia we have a form which, as far as our discoveries have progressed in reference to the Puerco fauna, may be considered ancestral to all the Wahsatch Condylarths and Perissodactyls. Dr. Schlosser* has gone further and singled this out as the so/e ancestor in the Puerco of the true Equine line, but in such an early geological period it is not possible to deter- mine whether Lwprofogonia may not alsc have been the ancestor of the Phenacodontide. We may suppose that the feet of uprotogonia were semi- plantigrade and provided with five well-developed digits, the elements of the podium being serially arranged ; while turning to the Wahsatch Hyracotherium we find quite a modernized arrange- ment of the podials, very different from that in Phenacodus. In Hyracotherium the first digit of the manus has disappeared, and the lunar has a broad articulation with the unciform, the scaphoid extending also on the magnum. Unfortunately we know nothing of the skeleton in Luwproto- gonta, and must depend upon the characters of the teeth. Com- paring the teeth of /uprotogonia with those of the Wahsatch suc- cessors, we find them, as is well known, much less specialized than in either of the Wahsatch Perissodactyla (//yracotherium, Sys- temodon). In Phenacodus, as well as in Luprotogonia, the second superior premolar has a simple external lobe, while in Zyracothe- rium and Systemoden, especially the former, this tooth has two well-developed external lobes, and the third and fourth premolars are still more complex. 1 Proc. Geol. Soc., 1857, p. 54. * Stammesgeschichte der Hufthiere (Morph.-Jahrb., Bd. xii, 1887, p. 11). 1895.]| Osborn and Earle, Fossil Mammals of the Puerco. ©9 Genus Protogonodon Scot. Mioclenus Cork, in part. Superior true molars tritubercular without a hypocone. Both intermediate tubercles distinct, but tending to coalesce with protocone, forming an internal crescent. Last inferior premolar simple in structure, and showing only in some specimens an indication of a deuteroconid. Inferior true molars with trigonid not raised above talon, and with paraconid well marked. Lower jaw long and slender. The genus Profogonodon was established by Prof. W. B. Scott,’ to include the species of AZzoclenus called by Cope MW. pentacus. We consider the separation of this genus from the typical form of Mioclenus, viz., the M. turgidus, to be a decided advance in our knowledge of these early Eocene forms. Scott in his valuable paper “A Revision of the Creodonta,’ already referred to, places the genus Protogonodon among the Condylarthra, and says: “I think there can be no doubt that this genus is referable to the Phenacodontide.” ‘The discovery of a series of upper molars, which should probably be referred to Profogonodon, causes us to assign this genus a position nearer the line leading to Artiodac- tyla ( Zr igonolestes), than to that leading to the Perisodactyla and Condylarthra of the Wahsatch (Zuprotogonia). In a short notice in the ‘Naturalist,’ Earle* has given his reasons for placing Profogonodon as the probable condylarthrous ancestor in the Puerco of the Artiodactyla, and we believe it holds the same relationship to that group as Luprotogonta does to the Perissodactyla. We have temporarily included Pvo/ogo- nodon in the family Phenacodontide, but further knowledge of its structure will probably prove that it should be placed in a new family. The supposed upper molars of this genus are quite dif- ferent in structure from those of Huprotogonia. Protogonodon pentacus (Co/c). Mioclenus pentacus COPE. Superior true molars with a strong external cingulum ; internal cingulum complete on last superior molar; the latter as long transversely as the first. Inferior true molars with external cingulum. Hypoconulid of last lower molar small and not widely separated from entoconid. 1 Proc. Acad. Nat. Sci. Phila., 1892, p, 322. 2 American Naturalist, 1893, p. 377- 68 Bulletin American Museum of Natural History. [Vol. VI, Dentition—TVhe superior molars (No. 954), which we refer to this genus, were not found with the lower teeth, but their general character and size are exactly what we should expect to find in the upper molars of Pyrofogonodon. The form of the superior molars in this genus is short and broad; the external cusps are very low and widely separated. The external lobes of the upper molars resemble somewhat those of the Creodonts (.Sarcothraustes), but are much lower. ‘There is a prominent external cingulum on Fig. 20, Protogonodon pentacus. Leftlower jaw. Superior and external view. Natural size. all the molars, which extends completely across the external face of the teeth. ‘The intermediate tubercles are strongly marked, and they have spurs which run outwards and join the anterior and posterior cingula respectively. The protocone is large and placed at the internal end of the median valley, and at an equal distance between the outer lobes. The protocone, like the inter- mediate tubercles, has crests running outwards from it, and when the teeth are much worn these elements of the crown unite, and asa result there is formed a well-marked internal crescent. ‘The first superior molar is square in outline, with its internal cingulum in- complete ; the second has the cingulum more developed than in the first, and in the third it is complete. In strong contrast to some of the other Puerco forms, the upper molars of Profogonodon can be said to be without hypocone; only on the first and second molars is there any rudiment of the hypocone upon the posterior cingu- 1895.| Osborn and Earle, Fossil Mammals of the Puerco. 69 lum. ‘This character of the upper molars of this species is very different from most of the Periptychide, where the hypocone is _ generally well developed. The last upper true molar in /. pentacus is as long transversely as the first ; it has two well- marked external cones, and the intermediate tubercles are dis- tinct. Comparing then the upper true molars of Protogonodon with other allied forms from the Puerco we find that its simple tri- tubercular molars and the want of supplementary internal cones, sharply differentiates this genus from the other Condylarthra. We would suggest that the most similar form to Protogonodon, in the characters of the upper teeth,is the A/zoclenus turgidus. In the latter genus the internal supplementary cusps of the upper true molars are wanting, or only feebly developed on the first molar. The first lower premolarin the P. pentacus is single rooted and separated by a short interval from the Pm. 2. The crowns of all the lower premolars are very simple in structure, and the last premolar of the type specimen is without a deuteroconid. Scott’ has figured two specimens of the last inferior premolar in Prodo- gonodon from Cope’s collection, and in both of these teeth there is a minute deuteroconid. However, in comparison with £u/o- protogonia, or any of the Periptychide, the last inferior premolar in Profogon- odon is much simpler in structure. If we compare this tooth with that of Trigonolestes, we observe a_ close resemblance in their general form and structure. The type specimen of P7vo- togonodon pentacus agrees with that of mi 8 3 Pantolestes, iv. the last inferior premolar Fig. 21. Protogonodon pentacus. ; : Superior true molars. Crown view. lacking a deuteroconid. The presence Natural size. of this cusp on the last inferior pre- molar is of general occurrence in all the Phenacodontide, and in one of the contemporaries of Protogonodon, viz., Euprotogonta, it is as large as the paraconid. The inferior true molars in Protoegonodon are low and broad (Nos. 951, 954) ; the trigonid is not raised above the talon, and a well-marked paraconid is present on all the molars. The external 1“Eyolution of the Premolar Teeth in the Mammals.’ Proc. Acad. Nat. Sci. Phila., 1892, Pp. 247- 7° Bulletin American Museum of Natural History. |Vol. VU1.| cusps of the lower molars are not so crescentoid in section as in Trigonolestes ; these cusps tend to unite very early with the meta- conids. The anterior spur of the crescents is also less well developed than in 7yrigonolestes. The lower true molars in this genus resemble somewhat those of the Arctocyonidz, but in this family the hypocone is generally present and the upper molars “are more or less completely quadritubercular.”’ The lower jaw in P?. fentacus (No. 950) is much elongated and slender. ‘The portion of the horizontal ramus below the true molar series is of the same depth throughout, but below the last premolar becomes more slender and decreases rapidly in depth towards the symphysis. We hold that the form of the mandible in Protogonodon is one of the strongest points in relating it to the Artiodactyla, and like the jaw in that group, more especially the selenodont Artiodactyla, the portion of the horizontal ramus behind the dental series rises abruptly upwards, leaving a deep concavity below. ‘The jaw of 77igonolestes closely resembles that of Protogonodon. RELATIONSHIP OF PROTOGONODON. We now propose to review the characters of Protogonodon and give our reasons for assigning this genus a position in or near the line leading to the Artiodactyla. 1. The supposed superior true molars of Pvrotogonodon pentacus are tritubercular, and without the internal supplementary cusps (protostyle and hypocone) so characteristic of the Periptychidee. In this character Protogonodon agrees with the earliest known American Artiodactyle, viz., 77zgonolestes. ty . The inferior premolars of Pvofogonodon are simple in structure, and only on the last tooth of this series is there any indi- cation of a deuteroconid. ‘This is another character like that of Zrigonolestes, and decidedly different from the buno- dont Creodonta (Sarcothraustes). . The inferior true molars are quinquetubercular in structure, there being a well-developed paraconid. ‘The presence of this latter cusp probably proves that the type of lower molar found in Protegonodon should be associated with a superior molar, which is tritubercular. . The elongated and slender lower jaw of Protogonodon, espe- cially the marked shallowing anteriorly, is like that of Trigonolestes, and the selenodont Artiodactyla in general. ios) aS Article IIL.—FOSSIL MAMMALS OF THE UINTA BASIN. EXPEDITION OF 1894. By HENRY FAIRFIELD OSBORN. I.—InTRODUCTORY NOTES. The mammalian life of the upper Eocene of North America is clearly recorded in four old lake basins, the northern or ‘ Wind River,’ the west-central or “ Bridger,’ the east-central or‘ Washakie,’ and the southern or‘ Uinta.’ The American Museum parties have now explored each of these four basins in succession, concluding with the Uinta exploration, which is the basis of this report. This uppermost or latest of the Eocene lake sediments was made known by Marsh’ in 1870. A fuller exploration of the Uinta by a party under Scott and Speir in 1886 resulted in the memoir, ‘ The Mammalia of the Uinta Formation,’* published in 1889 by Scott and Osborn. The American Museum sent Mr. O. A. Peterson into the Uinta in the autumn of 1893, but owing to restrictions upon the Uncompahgre Indian Reservation he was obliged to return after having secured only a few fossils. In the late summer and autumn of 1894, however, aided by Major Randlett, of Fort Duchesne, and by a permit from the Secretary of the Interior, Mr. Peterson was far more successful. He secured for the Museum a complete geological section along the White River, a collection representing about 150 fossil mammals, many of which are new, because his search was mainly in older and lower levels than those previously explored. A preliminary study of this fauna leads to the following results : 1.—GENERAL RESULTS. 1. Beneath the true Uinta fauna is a distinct fauna transitional to the ‘Washakie’ and ‘ Bridger’ of the east- and west-central basins. This contains undoubted horned ancestors of the Titan- 1*QOn the Geology of the Eastern Uinta Mountains.’ Am. Jour. Sci., 1871, p. 191-198. 2 Trans. Am. Phil. Soc., May 17, 1889. [71] 72 Bulletin American Museum of Natural History. |Vol. VU, otheres, yet of an older type than Dzflacodon, because the pre- molar teeth are simple. With these forms are found surviving members of the distinctively Bridger types, such as Uzntathertum, also several forms which have hitherto only been found in Washa- kie, such as Achenodon. Still more surprising is the appearance upon this sub-Uinta level of species of Alotherium and Hyenodon, genera which. have been considered of a distinctively Lower Miocene age. Below this level is a still older fauna not yet fully explored, containing a number of typical Washakie forms, also a new type of large mammal, Sphenocelus, apparently hitherto not known. 2. Species of Zelmatotherium abound in the sediments of this sub-Uinta level, and confirm Earle’s prediction that this genus was ancestral to the ‘Titanotheres. No true Pa/e@osyops has thus far been found. Zelmatothertum cornutum isin one of the direct ancestral lines leading to the Titanotheres. It showsa flat cranium, very long nasals and small naso-frontal horns. It was anticipated by Z. vallidens of the Washakie, with horns in a still more rudi- mentary stage. 3. The smaller fauna of the basin hitherto recorded is increased by a new rodent related to Paramys, and a new Monkey related to Microsyops. 4. The full characters of Amynodon intermedius are given by the complete skeleton of a young individual obtained in the true Uinta. These results are so important, and give such large promise for the future, that the Museum has sent a party back into the Uinta for the third and more thorough exploration of 1895. 2.—GEOLOGY OF THE UINTA BASIN. Mr. O. A. Peterson contributes the following preliminary obser- vations upon the geology of the Basin: ‘“ The Uinta Basin of northeastern Utah is bounded to the north by the Yampa and Uinta Mountains, to the west by the Wahsatch Mountains, and to the south and east by the Tavaputs Plateau and Book Cliffs.. The 1895. | Osborn, Fossil Mammats of the Uinta Basin. 73 basin is drained through the centre by the Green River with its tributaries the White River of Utah on the east, and the Duchesne River on the west ; these enter almost opposite one another near the Indian Agency of Ouray. “ty. As we enter the southeastern border of the basin over- looking Book Cliffs, some thirty miles south of White River, the Wahsatch or Coryphodon beds are met with, resting upon the Cretaceous, but as I did not explore this I cannot add anything to what is known as to the relations between the Laramie and Wahsatch sediment at this point. The entire sedimentary mass in this basin dips northwestwardly at an angle of about 8°, and is observed unconformably resting upon the upturned edges of Laramie, especially along the northern border of the basin. “2. Conformably overlying the Wahsatch are the Green River Shales, identical in appearance with the corresponding series in the Bridger Basin, and attaining nearly the same thickness. As we cross the basin northwestwardly from Book Cliffs we reach the White River near the Colorado and Utah State line. Some forty miles west of its junction with Green River the White River cuts through the Tertiary rocks exposing cafions and vertical walls of sometimes 400 to 500 feet in thickness from the river bed to the top. Here we obtain fine stratigraphical sections. “3. Conformably overlying the Green River shales is a series of hard brown sandstones of the same character as sandstones which are found capping these shales north of the Uinta Mountains. Alternating with this brown sandstone are clay layers of a greenish- gray color, and this whole series reaches a thickness of about 800 feet. Specimens were found in the sandstone ledges of this series which represent true Bridger types. Overlying this series is a well-marked stratum of a light reddish color about 20 to 4o feet thick. This is especially noticed in the eastern part of the basin where the most satisfactory stratigraphical sections can be obtained. “4. The most important and faunally rich series of sediments in the Uinta basin immediately overlies and conformably succeeds the last mentioned reddish clay stratum. ‘These beds are about 74 Bulletin American Museum of Natural History. \|Vol. VU, 350 to 4oo feet thick, and are composed of coarse brown sand- stones with alternating clays. ‘The largest part of the vertebrate collection secured by the party is from this level, and is of great interest owing to its transitional relationship between the true Bridger and the Uinta fauna. “<. We now reach the ¢rue Uinta’ or Brown’s Park beds of a fine-grained soft material much the same in appearance as the characteristic Bad Lands of South Dakota, with the exception of the color which is of a brick red ; in fact, the reddish tinge holds good throughout the entire Uinta sediment. At a distance these beds present a ferruginous aspect, and are about 600 feet thick. This uppermost strata of the Uinta Basin has hitherto been reported’ as resting unconformably upon the underlying Bridger sediment, but no observable breaks were found to distinguish the true Uinta from the underlying Bridger sediment. So the writer found it necessary in collecting fossils to divide the beds over- lying the Green River shales into three different levels, which are here arranged alphabetically in ascending position : “Horizon C.—True Uinta beds 600 feet thick. Sandstones and clays brownish and reddish, ferruginous. The strata are sometimes evenly bedded and firm, but often irregular and friable, and present the characteristic Bad Land appearance where the erosion has been most complete. ‘This is the level in which the Vale College and the Princeton expeditions have made their explorations, and it contains the true Uinta fauna. “Horizon B.—300 feet thick. Soft coarse sandstones and clays. “ Horizon A.—8oo feet thick. Hard brown sandstones imme- diately overlying the Green River Shales.” 3.—THE THREE FauNAL LEVELS. ‘These excellent observations supply one of the most important links in the American lake faunal chain, namely that between the ! King, U. S. Geological Exploration of the goth Parallel, Map r. 2 Charles A. White, ‘On the Geology and Physiography of a Portion of Northwestern Colo- rado and Adjacent Parts of Utah and Wyoming.’ U.S. Geol. Survey, Ninth Annual Report, p- 690-1. 1895. | Osborn, Fosstl Mammals of the Uinta Basin. 75 Washakie and the Uinta. . The explorations of the present year, 1895, may modify these results, but it is certain we have now not only established a complete faunal transition from the Bridger and Washakie beds upon the one side, to the true Uinta level or Horizon C upon the other, but have demonstrated a closer con- nection between the fauna of this basin and that of the lowest White River Miocene. SUCCESSION OF SPECIES IN THE UINTA BASIN. (Museum Catalogue Numbers.) (Am. Mus. Exp., 1894.)—Mesonyx, 1505. Miacis uintensis, 1895. Diplacodon, 1853, 1853a, 1861-2. Amynodon in- Horizon C.—Uprer LEVEL. termedius, 1933; indet., 1934-5, 1506. : Isectolophus annectens, 1827-8, 1927. Diplacodon elatus beds. - Triplopus ? obliquidens, 1928. Epi- About 600 feet. hippus ? uintensis, 1930. Protoreodon and Leptotragulus, 1800-18, 1826-a. Brown and red sandstones and Incertz sedis, 1829, 1874. clays, ferruginous. (Princeton Exp., 1886.)—Mesonyx uin- tensis, Hyopsodus gracilis, Plesiarcto- mys sciuroides, Leptotragulus proavus, Protoreodon parvus, Diplacodon elatus, Amynodon advenus, Miacis vulpinus. Microsyops uintensis, 1899, 1900. Mia- cis uintensis, 1896. Mesonyx uintensis, 1892. ?M. obtusidens, 1891. ? Hye- nodon, 1893-4. Paramys_ uintensis, 1go1. Telmatotherium cornutum (skulls), 1845-52, 1837, 1868; (jaws), Telmatotherium cornutum beds. 1854-5, 1858-9. T. hyognathum, 1856. ‘T. diploconum, 1863, ? 1870-1; About 350 feet. (skeletons), ates 1860, 1869, 1872. Soft coarse sandstones and clays. Incertze sed.,1864—7. Amynodon, 1932, 1936, 1830. Helaletes guyotii, 18209. Epihippus, 1930. Achzenodon inso- lens, 1819, 1825. Elotherium uintense, 1820-24, 1826b,c. Uintatherium, 1884-1890. Horizon &.—MIDDLE LEVEL. Horizon A.—LOWER LEVEL. Telmatotherium megarhinum, 1500, ? 1864-5, 1876, 1877. Amynodon, 1878. Triplopus, 1879. Indet., 1501-4, About 800 feet. 1880. Uintatherium, 1881. Spheno- ccelus uintensis. Telmatotherium megarhinum beds, | Hard brown sandstone. GREEN RIVER SHALES. 76 Bulletin American Museum of Natural History. (Vol. VXI, C.—Upper level. True Uinta. Déplacodon elatus beds. ‘This is the level of the Princeton and probably the Marsh explora- tions. It is distinguished by the presence of three genera—Dyp- lacodon, Protoreodon, Leptotragulus—which have thus far not been found below. It contains, however, several species which have also been found in the middle level. It is apparently distinguished by the absence of Utntatherium. B.—Middle level. ‘Transitional. Zelmatothertum cornutum beds. This is arich faunal level, hitherto unknown. 7Ze/mato- therium cornutum is very abundant. ‘This is related to the White River or Lower Miocene by the presence of an ancient species of Elotherium, and probably of Hyenoden. It is related to the upper level, C, by similar species of AZesonyx, Amynodon and Eprhippus, but is distinguished from C by the presence of Utntathertum. It is related to the eastern or Washakie Basin by the presence of Telmatotherium hyognathum and Achenodon tnsolens, and appar- ently to the Bridger by Helaletes guyotit and Mesonyx obtusideny, both of which determinations are however somewhat doubtful. A.—Lower level. Base. Zelmatotherium megarhinum beds. This level has been comparatively little explored. It contains 7°. megarhinum, also found in the Washakie, besides Amynodon, Trt- plopus and Uintathertum. We have now to ascertain what type of Ucsntatherium existed as a contemporary of Ze/matotherium cornutum. Judging from the limbs, it was a very large animal, and will not improbably be found to belong to the Ucntatherium cornutum Cope, which was obtained at the summit of Haystack Mountain, or the very top of the Washakie beds. I]l.— PRELIMINARY DESCRIPTIVE REPORT. PRIMATES. ‘There are apparently numerous remains of Monkeys, Rodents and other small animals in a large block of sandstone, not yet worked out. We adda new type of A/zcrosyops. 1895.| Osborn, Fossil Mammals of the Uinta Basin. | Microsyops uintensis, sp. nov. Third premolar elevated and laterally compressed. Fourth premolar very small with three cusps in trigonid ; a very small and short talonid. First molar with paraconid. Second molar lacking paraconid. The only Primate hitherto found in the Uinta Basin is Hyopso- dus gracilis Marsh. The type of this new species (No. 1899) is a small jaw containing two premolars and two molars. It comes from the ‘ 7° corn- utum \evel,’ and is distinguished from the M. gracilis Leidy by the greater complica- tion and relatively reduced size of the fourth premolar. The submolariform structure of Pz, and the enlarged lateral pair of incisors, are the distinctive features ; _ 2 ? Fig. 1. Microsyops uinten- of this genus. There is also an isolated _ sis, type,No.1899. Lowerjaw, internal view ; superior view. lower molar, No. Igoo. One and a half natural size. CREODONTA. The Uinta Basin Creodonta thus far known are A/esonyx and Miacis. Weadd an apparently new form related to Yyenodon. Miacis uintensis, sp. nov. Fourth lower premolar with a high protocone bearing two cuspules upon the posterior slope, terminating in a talonid; no cingulum. The third lower molar either very small and single-fanged or wanting. The type lower jaw of this species (No. 1896) was found in #, close beneath the true Uinta level. It differs from M. vulpinus Scott in the structure of the fourth pre- molar, a tooth which in the latter species presents a com- plete cingulum and no cus- Fig. 2. Miacis uintensis. A. Type, No. 1806. External view of jaw. natural size. 2B, Fourth pules. The trigonid of Mi lower premolar, No. 1895. 78 Bulletin American Museum of Natural History. [Vol. VU, is broken off, the talonid is broad and elevated upon the outer side. Mz is a small tubercular with a complete but very much depressed trigonid and a narrow talonid. Mz3 is represented by a very small single alveolus. Another jaw from the true Uinta level (No. 1895) contains a fourth premolar, which presents the same characters as the above. ? Hyznodon. This genus is apparently represented by a jaw (No. 1893) from the middle or ‘ 7. cornutum level,’ in which the teeth are too poorly preserved to afford means of definition. It is, therefore, not taken as a type. It presents many similarities to the Wyeno- don paucidens jaw of the White River formation, especially in its Fig. 3. //y@nodon. Lower jaw, No. 1893, internal and external views. Natural size. very long stout symphysis and in its triple mental foramina. The specimen consists of a right mandible and a detached condyle. it contains the fang of a small lateral incisor and of a very stout canine. Behind this are two alveoli either belonging to two single- fanged teeth, Pt and P2—or to one tooth, bifanged Pz. The formula is therefore uncertain; it is either P, 4 or P, 3. The 1895. | Osborn, Fossil Mammals of the Uinta Basin. 79 missing third and fourth premolars were stoutly bifanged and of similar size. ‘The three molars were also apparently equal sized, narrow and bilobed as in Hyenodon. Another specimen (No. 1894) may also pertain to this species. It consists of fragments of a molar tooth and of the limbs. Mesonyx Cope. There are two individuals belonging to this genus, both from the * 7. cornutum level.’ The first is a smaller animal (No. 1891) represented by the lower jaws and hind limb with a perfect foot, corresponding nearly in size with the J/. odbtusidens Cope, so fully described by Scott. The second is a very large skull apparently related to the following species : Mesonyx uintensis S. & O. This powerful mesoplacental is represented by a skull (No. 1892) belonging to a slightly smaller individual than the Prince- ton type, which was founded upon a series of lower molar teeth. Measurements. Total length of skull, estimated........ .44. mm. Incisors to condyles, . ep ns ASP Length premolar-molar series........... .137 Width across zygomatic arches.......... A ee The following are the principal characters: The cranium is slightly longer than the face. There is a high narrow occiput, extending forwards into a thin sagittal crest above a small brain- case. The frontals widen suddenly into a broad supraorbital plate which overhangs the rather small orbits. The nasals are long and narrow, but widen in the median line between the orbits. The maxillaries are compressed behind the canine, and perforated by an infraorbital foramen above M+. The lachrymals are widely exposed upon the face. The zygomata are slender, but arch widely outwards and then suddenly descend into the glenoid fossa, which is very deep and presents sharp pre- and post-glenoid crests. Behind the glenoid region the skull is very short and the paroccipital plate is relatively narrow. 80 Bulletin American Museum of Natural History. [Vol. VU, Fig. 4. Mesonyx uintensis. No. 1892. Base of skull. One-quarter natural size. Foramina.—There is a strong mastoid foramen. The for. ovale pierces the ridge between the glenoid fossa and the pterygoid border. ‘The periotic fills in the auditory meatus inferiorly, com- pressing the for. lac. medius and f. |. posterius into small spaces, behind which are the small condylar foramina. The most distinctive feature of the skull is the backward exten- sion of the posterior nares and the inclosure of the roof of the pharynx by two long palato-pterygoid plates, the lower borders of which incline towards each other in the median line so as almost to come in contact. This is paralleled by the well-known inclo- sure of the same region in certain species of //venodon. 1895. | Osborn, Fossil Mammats of the Uinta Basin. SI We are also struck by the many points of parallelism which this skull presents with that of Alothertum uintense, a totally unrelated form. These parallelisms are undoubtedly attributable to adap- tive conformation in both cases to a type of dentition and a mode of mastication which have many points in common. RODENTIA. This order is represented bya large number of remains of jaws and skulls contained in a block (Nos. 1907-1919) which has not yet been worked out. There are also the small undetermined jaw (No. 1908), the larger jaw (No. 1906) corresponding in size with Plesiarctomys sciurotdes, and a number of upper and lower teeth which probably belong to Paramys Leidy. Paramys uintensis, sp. nov. Upper molars quinquetubercular with posterior cingulum and a mesostyle. Lower molars quadritubercular. Crown crenulate. The type (No. 1tgor) is from the ‘ 7. cornutum level,’ and deserves description. The upper molars are strictly tritubercular with slight anterior and prominent posterior basal cingula; the three primary cusps (paracone, metacone and protocone) are prominent, as are also the two intermediates (protoconule and metaconule). ‘The lower teeth & are much larger than in Leidy’s types of P. delicatus, P. delicatior or P. delicatissimus. Marsh’s P. robustus' is still indeterminate, not having been adequately described or figured, although proposed twenty-three years ago. Theseteeth present asimi- Fig.s. Paramys J uintensis. Vy pe, lar, irregularly quadritubercular crenulated crown, No. :o01. Upper and lower molars, whereas in Plestarctomys we observe a smooth low- crown views. * apt Ae : _ Twice natural crowned type. /. winfenszs is further distinguished _ size. by the apparent absence of a paraconid. The upper molars furnish another link in the chain of evidence that the ancestors of the Rodents were tritubercular. 1 Am. Journ. Sci., Sept., 1872. [Afay, 1895.) 6 82 Bulletin American Museum of Natural History. |Vol. V1I, AMBLYPODA Cofe. Uintatherium Zerdy. The discovery of remains of this genus in the Uinta Basin is an important one. In Horizon A was found the head of a humerus (No. 1881). Inthe more fully explored Horizon 4 remains of seven individuals were found, as follows: Occipital region of a skull (No. 1884), a humerus (1885), two femora (1880-87), frontal horns of two individuals (1889-89a), miscellaneous footbones of several individuals (1890). We look forward with interest to the discovery of a skull from this level. PERISSODACTYLA. - TITANOTHERIID 2. PALASOSYOPINAL. One of the chief results of this expedition is the clearing up of the cranial and dental characters and of the systematic position of Telmatothertum, a work which has been so ably begun in Earle’s Memoir." Numerous remains of the skeleton were also procured, but the description of these is reserved for a subse- quent paper. fHTistorical Notes.—The following characters were assigned to Telmatotherium validum by Marsh’* in 1872: 1. Premaxillaries compressed with an elongated median suture. Zygomatic arch slender. Upper molars with inner cones elevated and pointed, and with a well-developed basal ridge. Upper canines large, pointed, with strong cutting edges. Incisors with inner basal ridge. Palate deeply excavated between the premolars. Nasals decurved laterally and much compressed. Last upper molar with a single internal cone. Diameter upper premolar-molar series, 224 mm. Type species, 7'e/matotherium validus Marsh. Specimen found at Henry’s Fork in the Main Bridger Basin. Date, July 22; separata, August I, 1872. 1*A Memoir upon the Genus Palzosyops Leidy and its Allies,’ Journ. Acad. Nat. Sci. Phila., Vol. IX. 2 Am. Journ. Sci. and Arts, Aug., 1872. 1895.| Osborn, Fosstl Mammals of the Uinta Basin. 83 Four species have been subsequently described : 2. The second species 7. (Paleosyops) vallidens Cope! was named by Cope September Ig, 1872, from a series of upper premolars and molars found in the Bitter Creek region or Washakie Basin. It was, however, identified with Pa/zo- syops and distinguished from P. major as follows: Molar teeth larger. Superior molars with two transverse ridges connecting the inner tubercle (protocone) with the outer crescents (paracone and metacone) enclosing a pit between them. Premolars with outer crescents fused into a single ridge. Summits of all the crescents elevated. All the teeth with strong internal basal cingula which rise up on the inner tubercle (protocone). Diameter upper premolar-molar series, 220 mm. 3. The third species described was the 7. (Leurocephalus) cultridens of Scott and Osborn? in 1878. The type specimen is an upper jaw with a com- plete set of teeth and part of a lower jaw with the grinding teeth. Also the posterior portion of one of the nasals and a part of the frontals. Diameter upper premoilar-molar series, Ig0 mm. The authors distinguished this type clearly from Palgosyops but not from TVelmatotherium, with which Earle has shown it to be identical, although specifically distinct from 7. validum. The locality is Henry’s Fork, Bridger Basin. 4. The fourth species, 7. (Palzosyops) hyognathum, was established by Scott and Osborn* in 1889, upon a very large jaw found in the Bitter Creek or Washakie Basin. It was characterized by its close series of procumbent incisors ; a symphysis extremely long and shallow; the canines rather small and semiprocumbent ; diameter of lower molar-premolar series, 245 mm; a large inferior diastema. Evidently related to Diplacodon. 5. The fifth species, 7. (Palzosyops) megarhinum, was proposed by Earle? in 1891 upon a fine skull, also from the Washakie Basin (Princeton Mus., No. 10,008). The teeth in this type are badly damaged, so that new skull charac- ters only could be assigned, namely : Molar with a shelf-like suborbital process ; face very short ; nasals very long, expanded distally ; premaxillary symphysis short and narrow ; palate narrow and arched. Diameter premolar-molar series, 148mm. No superior diastema. + 6. The new species 7. diploconum and 7. cornutum are here proposed. The type of 7. validum Marsh has not yet been figured, and its specific characters are still indefinitely known. ‘The Museum collection now contains specimens which we refer to 7. wval/idens, T. hyognathum and T. megarhinum, besides the new species 7. © adiploconum and 7. cornutum. 1 Proc. Am. Phil. Soc., Sept. 19, 1872, p. 487. 2 Bull. E. M. Museum Geol. & Arch., 1878, p. 42. 3 * Mammalia of the Uinta Formation.’ Trans. Am. Phil. Soc., 1889, p. 513- 4 Am. Nat., Jan., 1891, p- 46. 84 Bulletin American Museum of Natural History. (Vol. VU, Telmatotherium JZarsh. A genus partly contemporary with Pa/@osyops, but transitional in evolution to Diplacodon. An incipient fronto-nasal horn in the latest species. Nasals long and decurved laterally. Premolars simpler than molars. Upper molars with high pointed cusps, paracone and metacone approximated to protocone ; conules reduced or wanting. Comparative Measurements. Molar-premolar| Total length Width zygo- series. skull. matic arches. Telmatotheriumvalidum ... ...... .224 - vallidens, type..... .220 Ut Mh No. 1569. .184 .500 133) ay cultridens) ees. .185 = hyognathum, lower. -239 ~ megarhinum, type. . -145 re =) Nomis 002 -147 355 14 ss diploconum ....... -174 ?.450 ? 216 oa Commubum™sse meee .208 -565 -243 Telmatotherium megarhinum Zar/e. Superior premolar series, 148. No diastema. A broad suborbital shelf. A long narrow sagittal crest. When Earle established this species he left its generic position open, owing to the fractured condition of the teeth. The very complete skull (No. 1500) procured in 1893 from the lowest level, or Hor. A, contains perfect teeth, which are of the Ze/matotherium type; this fact, together with the presence of an infraorbital shelf, as in the 7. cornutum type, determine the generic position of this species. But this species with its long, thin and high sagittal crest presents a far more primitive condition than 7. cornutum. It differs from 7. cultridens also in the small oval section and short enamel area upon the canines, as well as in the infraorbital shelf and the posterior position of the infraorbital foramen. Its general characters are as folllows : Superior dentition : The incisors are small, nearly continous, with a rounded anterior con- tour. The canines are small and rounded. The premolars 2-4 present large single internal cones. ‘The molars exhibit basal cingula at the bases of the para- and metacones. There is a small hypocone upon M3. 1895.| Osborn, Fosstl Mammats of the Uinta Basin. 85 The malars present a thin shelf. The zygomata diverge slightly posteriorly. ‘The brain-case is small, and the sagittal and occipi- tal crests are very prominent ; the latter spread superiorly. The orbits are small and sunken, bounded by hook-shaped postorbital processes upon the frontals. The posterior nares open opposite the second molar. The postglenoid process is small. The pre- maxillary symphysis is apparently long. The infraorbital fora- men is just at the anterior margin of the malar, and the space between the orbit and the nasal notch is very short. We also refer Nos. 1864-5 to this species and numerous skeletal remains. Telmatotherium diploconum, sp. nov. Superior premolar-molar series, 174 mm. A large hypocone upon last upper molar. Naso-frontals without horn. Long sagittal crest. Canines small, rounded. ‘The type is a skull (No. 1863) in which the nasals are wanting and the mid-region of the cranium was crushed. This type is remarkable in being of the same stage of evolution as 7. cultri- dens, and yet occurring in the same level as the far more modern- ized 7. cornutum. This species differs from 7. megarhinum in the absence of the infraorbital shelf, and in the presence of a large hypocone upon the last upper molar. The premolar-molar dentition is similar in size and form to that of 7. cu/tridens, but there are the following important general differences : (1) Canines small and circular in section ; (2) a very short diastema, if any, behind the canine; (3) a large hypocone upon M2; (4) the infraorbital foramen close beneath the anterior border of the molar. In several respects it distantly resembles 7. megarhinum. The occiput is small and high. There is a long sharp sagittal crest extending about halfway forward to the orbits. The posterior nares 1s opposite the second molar. The premaxillary symphysis is relatively short. The zygomata are nearly parallel with the side of the skull, but they arch upwards posteriorly; there is a very prominent postorbital process. The form of the molar, at its junction with the pre- maxillaries, is very similar to that in 7. cultridens, but the infra- 86 Bulletin American Museum of Natural History. \Vol. VU, Fig. 6. Telmatotherium diploconum. Type, No. 1863. Superior and lateral views of skull. One-fourth natural size. The nasals are broken off. orbital foramen is immediately in front of the suture, instead of being placed well forwards. The incisors are wanting. The canines are very small and circular in section at the base. The second premolars are ina slightly more advanced stage of evolu- tion than in 7. cu/tridens. ‘The outer lobes of the molars are very much elevated, with feeble basal cingula. The hypocone of M* is quite as large as those of M+ or M2, 1895.| Osborn, Fossil Mammats of the Uinta Basin. 87 Telmatotherium hyognathum 5S. & O. Incisors, §. Inferior premolar-molar series, 224 mm.; lower molar cusps high and pointed. Three lower incisors. Lower canines rounded, followed by a wide diastema. This species is represented by a very large pair of lower jaws (No. 1856) containing ¢4ree incisors and agreeing in all other respects with the type of 7. Ayognathum. It is noteworthy that the canines have short enamel crowns and are formed like those of the Miocene Titanotheres, but the jaw itself differs widely from the Titanothere type in the long shallow symphysis. Telmatotherium vallidens Coe. Superior premolar-molars series, 184-220 mm. A narrow diastema. Molar cusps less elevated. A rudimentary naso-frontal tuberosity. Premaxillary symphysis short. Topof cranium flattened; very short bifid sagittal crest. The examples of this species are not from the Uinta but from the Washakie Basin, and were found by the Bridger expedition of 1893 in a brown layer of sandstone three miles north of the base of Haystack Mountain upon Bitter Creek. They are described here as an important link in the Telmatotherium series. The molar teeth as displayed in the two skulls (Nos. 1569, 1570) agree closely in every detail with those of Cope’s type’ also found in this basin, although they are considerably smaller, measuring only 184 mm. as against 220 mm. They are distinguished from the molars of Z. val/idum (Marsh) by the lower and more obtuse cusps. The skull is distinguished by the short premaxillary sym- physis. The species is distinguished from 7. hyognathum by the very narrow post-canine diastema; from 7. cultridens by more obtuse molars and by the naso-frontal tuberosity ; from 7° cornu- tum by the short posterior constriction of the temporal ridges into a bifid sagittal crest. When these skulls were discovered they were described by Dr. Wortman, in a letter from the field, as Wanteoceras or ‘ prophet- horned.’ But in the Museum great doubts were expressed by 1 See Tertiary Vertebrata, Plate 51, Fig. 1. 2 It appears as if,;Cope’s measurements were erroneous. 88 Bulletin American Museum of Natural History. (Vol. VU, Professor Cope and by others who examined them as to whether the tuberosities (47) above the orbits could really be regarded as in- cipient horns. These doubts have now been removed by the discovery of Z. cornutum,and Dr. Wortman’s observation is verified. Fig. 7. Telmatotherium vallidens. Composition, Nos. 1569-70. Side view of skull. One- fourth natural size. T. vallidens presents the first transitional features towards 7. cornutum and the later Titanotheres. The horns exhibit the most rudimentary stage imaginable ; they are borne more upon the frontals and less upon the nasals than in 7. cornutum. The cranium is broad upon the upper surface between the orbits and narrows very gradually towards the occipital region, where the two temporal crests converge. They do not, however, unite into a single sagittal crest, but leave a deep median pit followed by a narrow valley which opens out into a triangular space between the occipital crests. The occiput is very broad and low. In addition to those above mentioned, there are other features which separate 7. vallidens from 7. cornutum, especially (1) the absence of an infraorbital shelf upon the malar, (2) the short wide- spreading and relatively heavy form of zygomatic arches, (3) the relative shortness of the nasals, (4) the more slender postglenoid processes, (5) the shortness of the premaxillary symphysis. In general the face and the nasals are relatively shorter in 7. vad/i- dens than in 7. cornutum ; there is little or no diastema behind the canines ; the posterior nares open much further forwards, or ,- 2 > 1895.| Osborn, Fosstl Mammals of the Uinta Basin. 89 Fig. 8. YVelmatother7um vallidens. No. 1569. Superior view of skull. One-fourth natural size. between the second molars; there is no trace of a hypocone upon M3. The canine crowns are wanting; they are rounder in section and much more powerful than in 7. cornutum. The opposite molar series are not so nearly parallel. The true molars exhibit the Telmatotherium type, but it is less sharply defined than in 7. cultridens or in 7. cornutum. 90 Bulletin American Museum of Natural History. |Vol. VU, v Soy ANY iY ae ff y); y 4 Lvl SHG Ana] edn Fig. 9. TVelmatotherium vallidens. No. 1570. Superior view of skull. The upper surface of the frontals and nasals is abraded. There are only a few resemblances to Z. cornutum, such as the proportion of the teeth and the development of basal cingula, the reduction of the intermediate tubercles upon the molars. The general conformation of the zygomatic arches presents an affinity to that of the Zvtanotherium bucco type of the Miocene. Telmatotherium cornutum, sp. nov. Incisors 3. Premolar-molar series, 208 mm. A narrow diastema, Upper canines lanceolate. Long premaxillary symphysis. A well-developed naso- frontal protuberance. ‘Top of cranium completely flattened. No sagittal crest. An infraorbital process upon malar, The type of this species is a fine skull (No. 1851), while several other well-preserved skulls from the same levels give us all the 1895.| Osborn, Fossil Mammals of the Uinta Basin. gi =e fn -> SE —=— = SSS SSS ——= —- = SSSI EZ ——S—S—S— ————— = SS ——s Ta ~- = = —== =— = Zane = Z AD. oD SS MyaibyneDS DASNY aya Tau ir hd 4 ‘tt 5) a, i! l oom esl M/s Fig. 10. Telmatotherium cornutum. Type, No. One-fourth natural size. Fronto-nasal horn at H, 1851. Superior and side views of skull. al 92 Bulletin American Museum of Natural History. (Vol. VU, cranial characters and the superior dentition (Nos. 1850, 1847, 1848, 1852, 1837). Unfortunately none of these skulls have the jaws associated with them, but several more or less perfect jaws, although found apart, agree perfectly in size (Nos. 1857, 1858, 1854, 1855); they are all readily distinguished from the jaw of 7. hyognathum by the presence of only two incisors. This species is remarkable for its very long flat-topped cranium and its incipient knob-like osseous horns borne chiefly upon the nasals but partly upon the frontals. ‘These horns project later- ally and rise slightly above the general surface, and are best seen in the anterior view, Fig. 10. These characters and the absence of the fronto-parietal and interparietal sutures all point well towards Tztanotherium, but the premolars are still absolutely simple, showing no trace of the postero-internal cusps which char- acterize Diplacodon elatus. Other striking peculiarities are the upward arching Ape cranial region, the extremely long, narrow and laterally decurved nasals; the strong infraorbital shelf upon the molars (seen also in 7. megarhinum), the slender zygomatic arch, the low occiput, the backward extension of the posterior nares by the palatines, and the partial inclosing of the roof of the pharynx by the pterygoids. More in detail (No. 1851) the zasa/s almost overhang the pre- maxillaries, they are laterally compressed above the infraorbital foramina so as to give the impression of distal expansion ; the median fronto-nasal suture extends back beyond the mid-orbital line, but laterally the nasals terminate just above the orbits so as to include most of the incipient horn. The premaxzl/ary symphy- sis is elongate as in 7. validum. ‘The maxillaries are shut off by the very narrow lachrymals from the anterior border of the orbits. The infraorbital foramen is placed above M+ in front of the molar suture. The molars extend sharply upon the side of the face and then dip into the outwardly projecting shelf ; with an obtuse postorbital knob. The /vonta/s exhibit a prominent post- orbital hook ; there is a delicate lateral ridge marking the limits of the temporal fossa ; between these ridges the cranium is arched both from side to side and antero-posteriorly, presenting a very different form from the concave profile of even the oldest known Titanothere ; there is a slight constriction in the posterior third, 1895.| Osborn, Fossil Mammals of the Uinta Basin. 93 but the cranium is even here two inches wide, and there is not the semblance of the crest seen in 7. va//idens; the entire absence of the upper cranial sutures even in the young individuals (No. 1847) is a noteworthy Titanothere character. Owing to the sudden dipping of the superior contour the occiput is rather low and subquadrate in outline. In side view the faint tem- poral ridges can be traced to the superior angle of the occi- put. The zygomatic arch is very slender; it arches slightly upwards and very much less strongly outwards than in 7. vallidens. The postglenoid process is very thick in antero- posterior section. In palatal view we observe a diastema between the median Fig. 11. Tedmatotherium cornutum. Type. incisors and a post-canine dias- Ep yeaeyee sy ae tema of 28 mm. The molar series are placed closely parallel so that the palate is long, narrow and deeply arched, and the posterior nares opens far back behind the last molar. The deep and long pterygoids arch towards each other in the median line, forming a deep fossa. Foramina.—The alisphenoid canal is very long ; the for. ovale is widely separated from the for. lac. medium ; the for. lac. medium and the for. lac. posterius are very small and partly con- fluent ; the condylar foramen is midway between the condyles and the for. lac. medium. Lower Jaw.—The most perfect of the lower jaws is No. 1857 ; it agrees in size exactly with the type skull No. 1851. In propor- tion it is rather shallow and slender, but presents somewhat more angulation of the chin than in 7. Ayognathus. The most distinc- tive character is the extremely long hook-shaped coronoid process which extends back over the condyle. The symphysis is long and rather shallow. Dentition.—Inferior : A very distinctive and progressive feature is the presence of but two incisors in the lower jaw. The formula 94 Bulletin American Museum of Natural History. |Vol. VII, is thus 13, C+, P4, M3. A second Titanothere feature is seen in the relatively short, rounded canines of the lower jaw, which present a wide contrast with the compressed lance-shaped tusks of 7. validum and 7. cultridens; an especial feature is the absence of enamel upon the fang. It is to be noted, however, that the specific reference of these jaws is not certain. Superior: The incisor series of the type (No. 1851) present a third circle, but the median incisors are separated by a slight space; they all exhibit prominent posterior basal cingula ; the lateral incisor is considerably enlarged. The canznes have short, outwardly and forwardly directed but slightly incurved crowns, with rather sharp borders, a suboval section and posterior basal cingula. Behind a short diastema is the first premolar, a simple, conical crown with an internal basal ridge; the second, third and fourth premolars exhibit s¢zgZe blunt or rounded internal cones, incomplete cingula, a strong antero-external (parastyle) and a feebler postero-internal (metastyle) ridge. The molars have the generic conformation ; the third molar is the largest of the series, and exhibits a strong parastyle and mesostyle and a feebler meta- style ; there is a strong cingulum at the outer base of the para- cone, and a feebler one at the outer base of the metacone; the hypocone is feebly developed upon M3. All these teeth are well- worn, and the animal was fully adult. The superior dentition of No. 1850 belongs to a younger ani- mal with sharply defined characters. Here we see more plainly the resemblances to the type of 7. cu/tridens. The canines are laniariform, with sharp lateral edges, basal cingula less marked and enamel continued far down. The outer faces of the pre- molars and molars are prominent and closely approximated to the internal cusps. We observe also a trace of the paraconule upon M2, and a distinct paraconule upon M®. In this specimen the pterygoids are long and not so deep. Telmatotherium validum Jd/ars/. Superior premolar-molar series, 224 mm. Molar cusps high and pointed with rudimentary intermediate tubercles ; last upper molar without hypocone. Second premolar with strong internal lobe. Premaxillary symphysis long? No infraorbital shelf. 1895.| Osborn, Fosstl Mammals of the Uinta Basin. 95 For the sake of completeness this definition is framed from Marsh's brief description. Telmatotherium cultridens S. & O. Superior premolar-molar series, Ig90 mm. Molars with nearly obsolete conules. Second premolar with feeble internal lobe. Last upper molar with- out hypocone. Canines laterally compressed. Naso-frontals without tuber- osity. Premaxillary symphysis long. No infraorbital shelf. There are several possibilities of error in the separation of these species, and these cannot be removed until 7. hyognathum and 7. validum are known both in the upper and lower dentition. AMYNODONTID. The independent position of this family has now been com- pletely established by the discovery in the Miocene of the com- plete skeleton’ of MZetamynodon showing four fully functional digits in the fore-foot. Additional characters of the family are brought to light by a second complete skeleton (No. 1933) found by Mr. Peterson in the true Uinta or upper level, Horizon C. The specific position of this animal is difficult to determine owing to the immature state of the dentition. It is provisionally referred to .4. intermedius 5S. & O. Amynodon intermedius S. & O. Dentition: I3, C}, DP$, M3. Upper canines suboval in section, inclined forwards. Four deciduous premolars in both jaws. Four permanent pre- molars in the upper jaw.? Lower canines erect, triangular. The skeleton is that of a half-grown animal. The epiphyses are detached from many of the limb bones and from all the ver- tebre. As in all fossils from the clay matrix of this level the bones are considerably crushed. The adult was a rather slenderly built, long-limbed animal, exceeding the largest Tapir in size. The manus was considerably longer than the pes, but we observe 1 Foss. Mamm. of the Lower Miocene. White River Beds, Bull. Am. Mus. Nat. Hist., July, 1894, p. 208. 96 Bulletin American Museum of Natural History. |Vol. VII, a very slight disparity in length between the tibia and the femur, the radius and the humerus. The skull is broad and flat, of entirely different proportions from the skull of 4. anxtéguus, which is high and narrow. It has the deep depressions in front of the orbits of A. zztermedius, but a much longer face and longer nasals than in Metamynodon. Measurements. Total length of skull, premaxillaries to occiput... .53. mm. Length of humerus from facet to facet.. .....-- S302! ss radius it % OC ayn aeisietae BSOSe un i: femur ag ee PO oh ie ai eed Oo Ole ae os tibia a te ae er me po seh O° < mavenqoril INNS a oageasncond sooo ce s103 iy oy MMEtAATSall MMe ever yvaisi sy (eee! | eels) evel pita FY Os-innominatum, fotalleneth]>.----5...- .-.-.- Aye tee Dentition—There are three complete upper and lower incisors; the latter are well preserved, and exhibit a posterior basal ridge and median column. The permanent lower canines are erect and triangular, while the upper are directed slightly forwards and are rounded upon the outer and flattened upon the inner surfaces. These teeth are only partially extruded, and it is, therefore, diffi- cult to compare their form with those of the type of A. zuterme- dius. The four deciduous premolars are still in place. The molar teeth agree closely with the A. ¢wtermedius types in form and measurement. Skull—The cranium is considerably crushed. The premax- illaries are short, slender superiorly and barely in contact with the nasals. The maxillaries are deeply concave in front of the malar antorbital bar. The zygomatic arch is comparatively slender, wide below the orbit and not very deep; there are two knob-like projections, also characteristic of dZetamynodon. ‘The external auditory meatus is open below, and the paroccipital process is very slender, and curves forwards inferiorly. The occiput is rather broad and low, overhanging the condyles. The base of the skull displays a rather broad, slightly concave palate, the posterior nares opening behind the second molar. The zygomatic arch spreads into a broad, flat triangular space around the glenoid fossa, the postglenoid crest being slightly everted. 1895. | Osborn, Fossil Mammals of the Uinta Basin. 97 The basi-cranial axis (basi-occipital and sphenoid, presphenoid) is a prominent ridge upon either side of which is a deep depres- sion containing the for. lac. medium and posterius. The for. ovale is widely separated from the for. lac. medium, while the condylar foramen is very close to the for. lac. posterius. The alisphenoid canal is apparently very short ; the mastoid foramen is very distinct. In the top view of the skull the nasals are relatively long; they are separated posteriorly by a median forward projection of the frontals. The brain-case is very large; the sagittal crest is low and sessile. The jaws resemble those of A. aztiguus in the very high con- dyles and the narrow recurved coronoids. Vertebre.—The characters of the vertebre have not yet been completely studied. ‘The centra are rather small. The anterior dorsals present moderately long, slender spines. Fore Limb.—The humerus is of the same total length as the femur, owing to its very broad and prominent great tuberosity, which is connected by a low ridge with the lesser tuberosity. The deltoid ridge is strong and rugose. ‘The internal and exter- nal condyles are subequal in size. ‘The ulna and radius are strikingly long and slender; the ulnar shaft has a deep tnhedral section ; the radius presents a shallow humeral facet. ‘The meta- podials of the manus are longer than those of the pes ; they are of high slender proportions ; the third metacarpal is slightly the longest, but the foot is functionally tetradactyl. The magnum is of the typical Rhinoceros form. Hind Limb.—The innominate bones are elongate and slender. The superior border of the ilium is not evenly rounded, but is excavated towards the sacral border, and elevated and evenly arched towards the external border. ‘The ischial border is straight, and the pubic border deeply incurved. The pubis is _ short and the ischium long. The obturator foramen is com- pleted by an internal bridge, which is greatly reduced or wanting in Metamynodon. ‘The acetabulum has a deep pit for the liga- mentum teres. [May, 7895.] NX 98 Bulletin American Museum of Natural History. (Vol. V1, The femur is distinguished by the antero-posterior diameter of the great trochanter, the small head with a large ligamentum teres pit, the antero-verted lesser trochanter, the prominent third trochanter slightly above the middle line, the moderately heavy shaft. The crus is somewhat shorter than the femur, but there is less disparity by far than in AZetamynodon. The fibula is long and very slender. The astragalus is broad with a short neck, The calcaneum is distinguished by the flattened transversely placed tuber, as in A7efamynodon. ‘The cuboid is broad and flat, with a narrow calcaneal contact. There is no metapodial dis- placement ; these elements are rather short and stout. EQUIDA, TAPIRIDA, HELALETID/. The smaller Perissodactyla are represented by remains of /7- hippus, Tsectolophus, and Helaletes. The former (No. 1931), from the true Uinta, is represented by two femora and part of an astra- galus, corresponding in size with the Zp7Aippus uintensis S. & O. These parts of the skeleton are actually smaller than in Cope’s type of Hyracotherium venticolum from the Wind River Beds, or base of the middle Eocene. ; Several imperfect jaws and teeth (No. 1927) correspond in size with /sectolophus annectens S.& O. Triplopus is represented by a lower jaw (No. 1928) containing the fourth premolar and first molar. From the ‘7Ze/matotherium cornutum \evel’ comes also a fine maxilla (No. 1929) of Heluletes guyotii containing Pa-M3. In this specimen the fourth premolar has two complete transverse crests, the metaloph being as elevated as the protoloph ; it is, therefore, in a slightly more advanced stage of evolution than the Princeton type. The animal was also about one-fourth larger. It has not been hitherto reported from the Uinta Basin. INCERTA@ SEDIs. Sphenocelus uintensis. This new genus is represented by the posterior portion of a skull, which is distinct from any cranium known to the writer. Its most distinctive feature is the presence of a pair of pits in the 1895.| Osborn, Fossil Mammals of the Uinta Basin. 99 Fig. 12. Sphenocelus uintensis. Vype. Base of skull. One-fourth natural size. floor of the skull upon either side of the narrow presphenoid. These pits were at first mistaken for the for. lac. media, but more careful investigation shows that they are roofed over by bone, and apparently do not communicate at all with the cranial cavity. The pit on the right side is perfectly preserved and clearly exhibits these characters. ‘The measurements of the pits are 42 mm. long, 14 mm. wide, and 2 mm. deep. The skull has a long narrow cranium surmounted posteriorly by a sagittal crest, which diverges anteriorly into two decidedly convex sagittal ridges. The occiput is rather broad, and below it are two widely set occipital condyles which are directed obliquely downwards and backwards. On either side of these the ex- occipitals extend down into obtuse paroccipital processes, which are closely joined to the post-tympanics. The external auditory meatus is open inferiorly. In front of this the postglenoid 100 Bulletin American Museum of Natural History. |Vol. VU, Fig. 13. Sphenocelus uintensis. Superior view of cranium. process faces somewhat inwards ; the glenoid facet is L-shaped, two narrow arms extending out upon the squamosal, and a broad arm descending upon the postglenoid. The distinctive feature of the zygoma is the presence of a deep depression just behind the lateral arm of the glenoid facet. Skull Measurements. Widthacross zygomatic arches) 92.5.2 ..2e esters 23) Dosis FLGIGht OMOCCIPUts s cise wins) niate = ee as) a TAZ ISTCACEH ON Mia) Git okt che lactated kel wyaysionv leper Eh tri h ae Breadthiot occipital condyles a ...eienst teins tiga S Basi-occipital to top of sagittal crest...........-. r4qq The foramina of the skull are related to those of the Perisso- dactyla, for there is a long alisphenoid canal, upon the outer side 1895.| Osborn, Fossil Mammals of the Uinta Basin. Io! Fig. 14. Sphenocelus uintensis. Occiput. of the anterior opening of which is the foramen. Just behind the posterior opening of the canal is the foramen ovale, and between these foramina are the two pits above mentioned. ‘This foramen is separated by a very wide plate of bone from the for. lac. me- dius, which is partly filled by the periotic mass. - Stee) Fig. 15. Sphenocelus uintensis. Side view of cranium. The distinctive features of the skull may therefore be summed up as follows : Deep paired pits in the alisphenoids, and orbito- sphenoids upon either side of the thin presphenoid ; a long ali- sphenoid canal; foramen ovale widely separated from for. lac. 102 Bulletin American Museum of Natural History. |Vol. VII, medium ; condyles very broad ; foramen magnum large ; occipi- tal crest extending anteriorly into a short sagittal crest with convex sagittal ridges ; skull apparently long and narrow. The relationships of this form are very difficult to determine. It has been compared with /aditherium, but without revealing any very close resemblances. The alisphenoid canal suggests that it is a Perissodactyle, and the form of the posterior portion of the skull is certainly very similar to that of Chalicotherium, but it lacks the robust tympanics observed in the European form, and exhibits the anomalous paired depressions in the roof the pharynx which so far as known to the writer are unique. An especial effort will be made to secure the teeth of this animal in order to elucidate this problem. ARTIODACTYLA. Elotherium uintense, sp. nov. Orbits open posteriorly. No inferior projections upon the malars. A pre- glenoid crest. Premolars 4 or 3. This species is named wntense to emphasize the surprising fact of its discovery in the Uinta Basin or ¢rue Eocene. It is even older than the period of the true Uinta beds, since it comes from the Zelmathertum cornutum level and éelow the typical Uinta or © Diplacodon elatus \evel. Measurements of Skull. E. uintense. E. mortoni. Length condyles to premaxillaries......... -43. mm. Wadthezy comapicrarches ameter tlle! =1=/eii: agit °° Tleioht oMocelputy aes ear esse crac aacs SDTAT es Front of orbit to condyles....... dusts eh 2 ae .185 mm. Molar-premolatSenes errr. t-)-r-1siy-eae er sini O° IMO atS? ts) sqare tore ue eis elas tea avelorameters Kop 00S Antero-posterior diameter of canines. ..... Rei ae Comparison with Leidy’s original specimens of 4. (Archeothe- rium) mortont from the White River (Am. Mus., Nos. 443-4) shows that the 4. wntense skull was one-fourth larger and much more robust. In #. mortoni the sagittal crest is thinner, the supra-orbital plates are narrower, the swelling for the brain upon 1895. | Osborn, Fossil Mammals of the Uinta Basin. 103 the outer surface of the cranium is more sharply defined, the face is relatively straighter. In fact, the Miocene type is smaller and less specialized than the Eocene, and the relations are the reverse of what we should have anticipated. " _Fig. 16. Elotherium uintense. Type. Lateral view of cranium. One-fourth natural size. It is readily distinguished from Achenodon robustus of the Washakie beds by the great elongation of the face and the short- ening of the cranium, both of which characters relate it to //othe- rium. It agrees with Achenodon and differs from the oldest types of Elotherium, however, in the orbits, which are widely open posteriorly. Unfortunately the premolar formula is uncertain, all the teeth are broken off and it is not possible to determine whether both P+ and P? were present; there was either one single-rooted followed by one two-rooted tooth, or there were two one-rooted teeth here. It seems most+probable that there were only three premolars. The three incisors increased in size laterally, the I* being much the largest. The canine tusks were very powerful. The molars, so far as preserved, resemble those of Achenodon robustus. The premaxillaries exhibit a wide contact with the nasals. The nasals are very long, narrow and indented anteriorly ; they extend posteriorly to a point opposite the middle of the orbits. The lachrymals are extensively exposed upon the face. The infra- orbital foramina are nearer the orbits than in Achenodon. ‘The frontals form a very broad, centrally depressed, plate over the 104 Bulletin American Museum of Natural History. |Vol. VMI, Fig. 17. Elotherium uintense. Superior view of cranium. orbits, overhanging but leaving them widely open; the orbits are much larger than in A. robustus. ‘The frontals converge sud- denly into the rather short, very sharp and high sagittal crest. The brain-case is very small. The occiput is high and expands fan-like superiorly as in H/othertum. ‘The zygomatic arches are slender in vertical diameter and lack the downward malar plates ; as shown in top view they diverge or arch sharply outwards. As in Achenodon the glenoid fossa is deeply depressed and there is a prominent preglenoid crest. ‘The palatal surface displays the posterior nares opening behind the last molar; the remains of a 1895.| Osborn, Fossil Mammals of the Uinta Basin. 105 deep pterygoid fossa ; a very broad glenoid fossa ; a short wedge- shaped basi-cranial series (basi-occipital and sphenoid, pre- sphenoid). The collection embraces a complete artiodactyl hind limb (No. 1820), including a femur, tibia, astragalus and calcaneum, cuboid and a metatarsal. The total length is 90 cm. or 35% inches. In comparison with that of “/otherium the femur is very short, and there were apparently four metatarsals, as indicated by facets upon the median pair. If this limb is related to the above skull it would distinguish it as a new generic type which might be named Protelotherium, characterized by four digits in the pes. Achznodon insolens Co/e. It is interesting to find this species, which is characteristic of the Washakie, upon the same level asthe E/otherium. It is repre- sented by a lower jaw (No. 1825) which corresponds closely with Cope’s type." The crown of the teeth are crushed and broken. There is also a portion of a lower jaw (No. 1819) con- taining the fourth deciduous premolar and an unworn first molar. The smaller selenodont Artiodactyla are represented by lower jaws and teeth provisionally referred to Leptotragulus and Proto- reodon (Nos. 1800-1818, 1826-18262), but they add nothing to cur knowledge of these selenodonts. 1 Tertiary Vertebrata, p. 433, pl. 57- mepiere 111.—ON THE SPECIES OF THE GENUS REITHRODONTOMYS. By J. A. ALLEN. INTRODUCTORY NOTE. The American Museum of Natural History has recently acquired a large number of specimens of the genus Rerthrodon- tomys. In attempting to determine them it was found necessary to consider the status of several obscurely known species, particu- larly the Rezthredon montanus and R. megalotis of Baird. The original purpose of the present paper was to settle, if possible, the character and relationships of these species, and to record several apparently new forms of the genus, the material at hand being too limited for a detailed revision of the group. The paper was originally prepared on these lines, and on the basis of the specimens belonging to the Museum collection. After the first draft was practically completed, Dr. C. Hart Merriam, Chief of the Division of Ornithology and Mammalogy of the United States Department of Agriculture, hearing of my work on the group, most generously and without solicitation, placed in my hands for use in this connection all of the United States specimens of this genus belonging to the collection of the Department of Agriculture, collected under his direction, and also those contained in his own collection. ‘These number alto- gether about 700 specimens, representing nearly the whole United States range of the genus, so that the total number of specimens available for study is not far from 925.’ This large amount of material throws much light upon the geographic distribution of the genus, and the manner of its representation over the diverse 1 Besides the Museum Collection and the specimens furnished by Dr. Merriam, I am indebted to Prof. L. L. Dyche, of the University of Kansas, for a series of nearly 40 specimens from the vicinity of Lawrence, Kansas, and to Mr. Gerrit S. Miller, Jr., of Peterboro, N. Y., for small series from central Kansas and northeastern Colorado. I amalso indebted to Mr. F. W. True, Curator of Mammals in the United States National Museum, for kindly sending me the type and only known specimen of Reithrodon montanus Baird, and also for other historic material mentioned fassz in the present paper. [107] 108 Bulletin American Museum of Natural History. [Vol. VU, climatic areas embraced within its range. It is, however, insuf- ficient for a final revision of the subject, so that the conclusions here presented must be considered as tentative, and the paper as merely a contribution toward a better knowledge of the group. HISTORICAL SUMMARY. The history of the group is, in brief, as follows: In the days of Audubon and Bachman the genus was known only from the vicinity of Charleston, South Carolina, and Liberty County, Georgia. The first species commonly referred to this genus was described by Audubon and Bachman in 1841, under the name Mus humulis (changed by them to Aumilis in 1851). Whether this species is correctly referable here, or is even certainly deter- minable, will be considered later. In the following year the same authors redescribed their A/us humulis, and added Mus lecontit and Mus carolinensis. The pertinency of JZus leconti to what is now recognized as Retthrodontomys 1s beyond question, and it is the first name that can be unequivocally applied to the south Atlantic coast form of the genus. Jus carolinensis has never been certainly identified, having proved a stumbling block to all subsequent writers on the group. The probabilities are that it was based on an immature example of Peromyscus’ (late Sztomys, late Vesperimus, = Hesperomys of earlier writers), probably P. leucopus’ gossypinus, and not at all referable to Aetthrodontomys. (See more at length on these points beyond.) In 1853 John Leconte referred AZus humulis and Mus carolin- ensis to Hesperomys, and ALus lecontti (for the first time) to the 1 Cf. oes! Ann. antl hee Nat. Gist. (6), XV, Feb A cane p. 192. 2 In June, 1894, I discussed (this Bulletin, III, pp. 294-7) the question of americanus Kerr (1792) vs. dewcopus Rafinesque (1818) raised previously by Dr. Coues, but left by him unsettled, owing to his inability to consult Kerr’s work. I was formerly familiar with Kerr’s work (Animal Kingdom, etc., 1792), and presumed that a transcript of Kerr’s description of his us agrarius americanus would decide all doubts in the matter. ‘The work not being in any library in New York City, I sent to a friend in Boston for an exact copy. of the passage in question. This settled beyond doubt the pertinency of Kerr’s name americanus to the White-footed Mouse of the northeastern United States, usually known previously as dewcopus Rafinesque, whereupon (1.c.) I adopted Kerr's name. Mr. Oldfield Thomas, on the authority of Mr. Gerrit S. Miller, Jr., has recently stated (Ann. and Mag. Nat. Hist., Feb. 1894, p. 192) thatthe name americanus is preoccupied by a Wus americanus occurring four pages earlier inthe same work. Through the kindness of my friend Mr. Samuel Henshaw, Secretary of the Boston Society of Natural History, | have in hand the copy of Kerr’s work belonging to the Society, from which it appears that Mr. Miller’s statement is well founded. ‘The M7us americanus ‘err (l.c., p. 227) is not identifiable, but probably relates primarily to some introduced species of Aus, ates conjectured by Kerr to be probably the Rat referred to by Kalm as living *‘among stones and clefts of rocks in the Blue Mountains of Virginia.’’ In any case the name americanus is untenable for any form of the White-footed Mouse. 1895. | Allen, Species of the Genus Reithrodontomys. 109g genus Retthrodon, at the same time claiming personal acquaintance with each. The next important reference to the group is by the late Pro- fessor Baird, who, in 1855, described a second species as Retthrodon montanus, based on a single specimen from the mountains of Colorado (exact locality unknown). In 1857 the same writer treated the group monographically, describing as new Retthrodron megalotis from near San Luis Springs, Sonora, and 2. longicauda from Petaluma, California, and recognizing four species as valid (the three described by himself and the old A7us Aumulis of Audu- bon and Bachman), and a fifth (2. carolinensis ex Aud. & Bach.) provisionally. In 1860 De Saussure added, from the mountains of Vera Cruz, Mexico, still another, under the name 7. mexicanus, and in 1861 described a second, under the name &. sumichrasti, also from Mexico. The next important original work on the group is Coues’s re- vision of the genus in 1874, and his more extended monograph of the group in 1877. The genus Rezthrodon is here shown to be exclusively South American, and for the North American species heretofore referred to Retthrodon be proposed the generic name Ochetodon. The species recognized by Coues in 1874 were (1) O. humilis (Aud. & Bach.), to which he referred 2. mega/otis Baird, and provisionally A/ws carolinensis Aud. & Bach. ; (2) O. longicauda (Baird); (3) O. mexicanus (De Sauss.), to which he referred provisionally specimens from Louisiana, thus for the first time recognizing this type of the genus as occurring in the United States. He also recognized provisionally (4) O. montanus (Baird), and (5) O. sumichrasti. In the later monograph the same alloca- tions are repeated, except that no reference is made, even in syn- onymy, to A. sumichrasti. In 1892 Merriam called attention to the fact that the name Ochetodon Coues was antedated by one year by the name Pezthro- dontomys Giglioli ; upon which showing this latter name quickly became current among North American mammalogists. In 1893 the present writer revived both #. mega/otis and RK. montanus of Baird, the latter, however, with some reservation, and gave the alternative name &. aztecus for specimens from 110 Bulletin American Museum of Natural History. |Vol. VII, northwestern New Mexico provisionally referred to &. megalotis. Later the same writer described as a new subspecies Rezthrodon- tomys mexicanus fulvescens from Oposura, Sonora. In 1893 Mr. Rhoads described as a new species Reithrodonto- mys pallidus from Santa Ysabel, San Diego Co., California. Excluding as of doubtful reference to this group both dZws carolinensis and Mus humulus of Audubon and Bachman (see beyond), the following nine species and subspecies have been described : 1841. Retthrodontomys lecontit (Aud. and Bach.). 1855. : montanus (Bd.). TOb7, i megalotis (Bd.). 1857. 3 Jongicauda (Bd.). 1860. : mexicanus (De Sauss.). 1861. i sumichrastt (De Sauss.). 1893. s aztecus Allen (provisional name). 1893. + pallidus Rhoads. 1894. - mexicanus fulvescens Allen. Prior to 1855 the group was known only from the coast region of South Carolina and Georgia. In this year Baird described a species from the “‘Rocky Mountains, lat. 38°,” and in 1857 extended the range of the genus to northern Sonora and Califor- nia, recording also specimens from St. Louis, Missouri. In 1860 (as above stated) a form was made known from the State of Vera Cruz, Mexico. In 1874 Coues referred to specimens from Louis- iana, Kansas, Iowa, Nebraska and Utah, and in 1877 gave the detailed records of his material, which included also localities in California and in southern Mexico additional to those mentioned by Baird. Alston, in 1880, recorded specimens from Coban and Duefias, in Guatemala. During the last two years the published addi- tional records include San Diego Co., California (Rhoads), Texas (Allen), Florida (Chapman and Rhoads) and Arizona (Allen). ‘To show the increase in material, as well as in our knowledge of the geographic range of the group, the following may be of interest. 1895. | Allen, Spectes of the Genus Reithrodontomys. Ii! In 1857 Baird’s material consisted of 32 specimens, represent- ing 7 localities ; 12 of the specimens were from South Carolina and Georgia, and 15 of the remaining twenty from the vicinity of San Francisco, California ; in other words, nearly all of Baird’s material came from two small areas on opposite sides of the con- tinent. In 1877 Coues recorded 57 specimens, representing 16 localities, the 25 specimens additional to those examined by Baird including 9 from southern Mexico, 3 from Louisiana, 4 from the coast region of central California, 6 from eastern Kansas, 2 from Utah, and 1 each from Iowa and Nebraska. MATERIAL EXAMINED. The material on which the present paper is based numbers 920 specimens, representing 166 localities, distributed about as follows : California, 87 localities and 471 specimens ; northern Lower California, 4 localities and 8 specimens ; Nevada, 7 locali- ties and 66 specimens ; Arizona, 3 localities and 25 specimens; northern Sonora, 2 localities and 5 specimens; Utah, 10 localities and 53 specimens ; New Mexico, 4 localities and 78 specimens ; Colorado, 3 localities and 15 specimens ; Nebraska, 8 iocalities and 27 specimens; Montana and South Dakota, 2 specimens each ; Kansas, 6 localities and 53 specimens ; Arkansas, 1 locality and 2 specimens; Louisiana, 3 localities and 13 specimens; Texas, 18 localities and 67 specimens; Tamaulipas, Mexico, 2 specimens ; Florida, 1 specimen ; Riceboro, Georgia, 6 speci- mens ; Raleigh, North Carolina, 61 specimens ; Southern Mexico, 2 localities and 2 specimens (Mazatlan and Tehuacan); Costa Rica, 1 locality and 17 specimens. The material in hand, while so extensive and covering such a wide range of country, is far from sufficient to properly repre- sent the genus throughout its range, large areas where it probably occurs being wholly unrepresented, while other portions of great extent are very inadequately represented, and only small sections of the general habitat with any great degree of fullness—mainly those areas covered by the Biological Surveys carried on by Dr. Merriam under the Department of Agriculture. 112 Bulletin American Museum of Natural History. (Vol. VU, GENERAL REMARKS. Geographical Distribution.—TVhe genus is not as yet known to occur in the Gulf States between Florida and Louisiana ; but this region has thus far been too imperfectly explored to render it safe to assume that it is absent from this coast belt, where the condi- tions are apparently highly favorable to its presence. Neither is it known to occur in the area to the northward between the coast region of the Carolinas and the Mississippi River. With this exception the genus is now known to have a practically continuous distribution from the coast of the Carolinas across the continent to the coast of California, and from the mouth of the Big Horn Riverin Montana southward to central Costa Rica, including both coasts of Mexico. From St. Louis, Missouri, westward to the Pacific coast the genus is apparently represented almost continu- ously, the higher altitudes in the mountains being of course ex- cepted. It also occurs across southern Texas, from about the mouth of the Pecos River eastward to the coast. List of Forms Recognised.—In the present paper fifteen forms are recognized, as given in the following list, which also states the number of specimens of each examined. 1. Reithrodontomys lecontit (Aud. & Bach.). Coast region of the South Atlantic States. Specimens examined, 69. 2. R. merriamt, sp. nov. Coast region of western Louisiana and eastern Texas. Specimens examined, 10. 3. R. dychei, sp. noy. Eastern Kansas and southeastern Nebraska, east to St. Louis, Mo. Specimens examined, 51. 4. R. dychet nebrascensis, subsp. nov. Colorado east of the Rocky Moun- tains, western Kansas, and north to southeastern Montana. Specimens examined, 43. ~) 5. R. montanus (Baird). Head of San Luis Valley, Colorado. Specimens examined, 1 (type of the species). os > . megalotis (Baird). Western New Mexico, eastern Arizona, and north to northern Utah. Specimens examined, 126. ~s = . megalotis deserti, subsp. nov. Death Valley region of southern Nevada and Inyo Co., California. Specimens examined, 189. 8. R. longicauda (Baird). Central California, west of the Sierra Nevada. Specimens examined, 175. 1895. | Allen, Spectes of the Genus Reithrodontomys. 113 g. R. longicauda pallidus (Rhoads). Southern California and northern Lower California. Specimens examined, 157. 10. &. arizonensis, sp. nov. Chiricahua Mountains, Arizona. Specimens ex- amined, 5. It. &. mexicanus (De Saussure). Southeastern Mexico. 12. R. mexicanus intermedius, subsp. nov. Valley of the Lower Rio Grande and adjoining coast region of Texas and Mexico. Specimens exam- ined, 36. 13. R. mexicanus aurantius, subsp. nov. Western Louisiana and eastern Texas. Specimens examined, 34. 14. R. fulvescens Allen. Northern Sonora. Specimens examined, 3. 15. &. costaricensis, sp. nov. Central Costa Rica. Specimens examined, 17. These forms present wide extremes as regards size, coloration, size and form of the ear, and ratio of tail length to total length ; but the connecting stages are so minutely graduated that none of these features, or any combinations of them, are serviceable as a basis for a sharp. division of the genus into minor groups. Nor do the cranial or dental characters prove any more satisfactory as a basis for minor divisions ; and no attempt is made in the present paper to make use cf them for the discrimination of species and subspecies, as on measuring a considerable series of skulls it soon becomes evident that the range of individual variation consider- ably overlaps the average differences between closely allied forms. The length of the skull varies, in different species, from 18 to 24 mm., but the conformation is practically the same in all the species. The average total length of the animal varies in the different species from about tro mm. to about 190 mm.; the extremes carry the total range from about 100 to 200 mm. _ In some species (as shown in the subjoined synopsis) the tail vertebrae form decidedly less than half (from 46 to 48 per cent.) of the total length ; in others they constitute more than half (from 52 to 58 per cent.) of the total length ; in others still the two measurements are practically equal, specimens from the same locality falling either side of the line. In the smaller, short-tailed South Atlantic and Gulf coast forms the general color above is dusky brown; the larger short-tailed interior forms are grayish brown with a tinge of fulvous; the [ Way, 7895.) S 114 Bulletin American Museum of Natural History. [Vol. VII, longer-tailed forms (ongicauda and mexicanus groups) are darker and more or less washed with bright fulvous, sometimes approach- ing golden rufous, while the largest and longest-tailed form of the group is nearly as golden rufous as the Golden Mouse (Peromyscus aureolus) of the South Atlantic States. In R. megalotis the ear is very large in comparison with most of the other species, the increase being not only in length, but more especially in breadth, and hence is quite different in form from the narrow and more pointed ear in A. lecontiz, R. merriamt, R. longicauda, etc. In R. dychei the ear is somewhat intermediate in size and form, making in &. dychet nebrascensis a decided approach to that of 2. megalotis. The following synopsis may aid in the determination of the species and subspecies, and serve to show, to some extent, their mutual relationships. Synopsts of the Spectes and Subspecies.’ A. Tail vertebrz less than half the total length. Ears small. a. Size small. Total length, 120; tail vertebra, 56; ear, 9.5. Above dark brown with a light wash of dark cinnamon brown, generally slightly darker along the median line ; below dingy gray, sometimes with a slight wash of yellowish ; lateral line usually indistinct or Above darker, prevailing color above dusky brown, with a prominent blackish median area ; sides yellowish gray brown, with an indistinct fulvous lateral line; below gray with a slight suffusion of yellowish DrOWile A ircinerseeb ters oh deine ee rom reeves sos. »ACUTET RUE c. Larger. Length, 130; tail vertebre, 60; ear, 10. Above fulvous gray lined with black, deeping on the sides to an indistinct fulvous lateral band >"below grayishewitttel je. sess ater R. dychei. d. Slightly larger, and more strongly suffused with fulvous. Length, nase whailevenrtebres OAs eae ell sims .. L. dychei nebrascensis. e. Very small. Length, 102; tail vertebrae, 51; ear, 10. Above pale yellowish gray brown, more yellowish on sides; below dull whitish. R. montanus. #. Tail vertebree about one-half the total length. Ears large. a. Size medium. Length, 136; tail vertebre, 63; ear, 12.5. Above yellowish gray, lined with darker ; lower border of sides more fulvous ; below white a0 202s fate as vcs ahaa shalane voter atari ... «+R. megalots. 4. Resembling the last, but with relatively longer tail. Length, 136; tail Vertebrae, 705 (ear. 2555 251). «an. een Seen R. megalotis deserti. 1 Measurements in millimeters. Unless otherwise stated, all measurements given in this paper are the collector’s measurements from fresh specimens, except those of the ear, which are always from the dry skins. ‘The measurement for the ear is the height from the notch, 1895. | Allen, Species of the Genus Reithrodontomys. I15 C. Tail vertebrz slightly more than half the total length. Ears smaller than in B a. Size medium. Length, 140; tail vertebree, 74; ear, Ir. Above yellow- ish brown, lined with blackish, with generally a darker median dorsal area ; sides brighter, the lower border forming a prominent bright fulvous lateral line ; below clear mie white, occasionally with a olin Deas ELE Accatk <1) ai Mi hme Veo Se oe R. longicauda. 6. Slightly larger than the fast ; cikecstinn paler. ..&. longicauda pallidus. c. Larger. Length, 150; tail vertebrz, 78. In coloration most resem- Vicia en ON LEC UUG ehh ata tia Du a ear ae ee R. arizonensis, D. Tail vertebree much more than half the total length. a. Large. Total length, 150; tail vertebre, 87; ear, 12.7. Above dull ferrugineous brown, becoming bright orange tawny on lower edge of sides ; below white with usually a yellowish cast..... R. mexicanus. 6. Larger and colors much paler.. Length, 176; tail vertebrze, gg ; ear, 12. Above grayish brown with a yellowish wash; sides strong yellowish fulvous ; below dull whitish. ...2. mexicanus intermedius. Re About the size of the last, or slightly smaller ; colors much stronger. Length, 168 ; tail, 94 ; ear, 11.5. Above strongly yellowish brown, with a blackish median area ; sides rich orange rufous ; below white with a faint yellowish tinge... ........... R. mexicanus aurantius. d. Rather larger than the preceding. Length, 176; tail vertebra, 100 ; ear, 11.5. Above pale yellowish gray, lined with black, with a blackish median area ; sides light yellowish ; below white. R. fulvescens. e. Largest ofthe genus. Length, 191; tail vertebra, 114; ear, 12. Above _ bright ferrugineous brown, finely lined with blackish, but with no distinctly darker median area ; sides orange rufous; below white, generaily with a slight tinge of yellow . ..... ...R. costaricensis. DESCRIPTIONS OF THE SPECIES AND SUBSPECIES. Genus Reithrodontomys Gig/io/7. Mus Avuv. & Bacu. (1841-51). Llesperomys WAGNER, Wiegm. Arch. 1843 (2), p. 51 (simply referring Audubon and Bachman’s “‘fiinf neue arten”’ of A/us to Hesperomys). Reithrodon \.ECONTE, Proc. Acad. Nat. Sci. Phila. 1853, pp. 410, 413 (merely refers AZus lecontii of Aud. and Bach. to Retthrodon ; not Retthrodon Waterhouse, 1837). : Reithrodor BAIRD, Mam. N. Am. 1857, p. 447 (not of Waterhouse). “* Reithrodontomys GIGLIOLI, Richer. intorn. alla Distrib. Geog. Gener. 1873, p. 60.” (Apud MERRIAM, Proc. Biol. Soc. Wash. VII, 1892, p. 26, footnote.) Ochetodon Cours, Proc. Acad. Nat. Sci. Phila. 1874, p. 184 (=Retthrodon Baird, ec Waterhouse. Reithrodontomys is the only North American genus of Muridz having grooved upper incisors. In other respects the cranial and dental characters are much as in Peromyscus. Externally the species also greatly resemble those of this latter genus, but are 116 Bulletin American Museum of Natural History. [Vol. VU, generally smaller, except 2. costaricensis, which has more resem- blance externally to some of the smaller species of Oxyzomys. I have to regret that Giglioli’s work wherein he established the genus Reithrodontomys is not accessible to me, and hence take the name on Dr. Merriam’s authority, as cited above. Reithrodontomys lecontii (Aud. & Bach.). LECONTE’s HARVEST MOUSE. ? Mus humulis Aup. & BAcH. Proc. Acad. Nat. Sci. Phiia. I, 1841, p. 97; Journ. Acad. Nat. Sci. Phila. VIII, 1842, p. 300. Vicinity of Charleston, S.C. (Not satisfactorily determinable ; probably not Retthrodontomys.) ? Mus humilis Aup. & BacH. Quad. N. Am. II, 1851, p. 103. (Habitat extended to vicinity of New York City.) ? H\esperomys| humilis LECONTE, Proc. Acad. Nat. Sci. Phila. 1853, p. 413 (in text). Reithrodon humilis BAIRD, Mam. N. Am. 1857, p. 448. Ochetodon humilis Cours, Proc. Acad. Nat. Sci. Phila. 1874, p. 185; Mon. N. Am. Roden. 1877, p. 123. (The Atlantic coast specimens and refer- ences only.) ?? Mus carolinensis AuD. & BACH. Journ. Acad. Nat. Sci. Phila. VIII, 1842, p. 306. ‘‘ Maritime districts of South Carolina.” (Not determinable ; probably a young Peromyscus.) 2? H|esperomys| carolinensis LECONTE, Proc. Acad. Nat. Sci. Phila. 1853, p. 413 (in text). ?? Reithrodon carolinensis BAIRD, Mam. N. Am. 1857, p. 452 (from Aud. & Bach.). Mus lecontii AuD. & Bacnw. Journ. Acad. Nat. Sci. Phila. VIII, 1842, p. 307. Georgia ; Ashapoo, S. C. Retthrodon lecontet LECONTE, Proc. Acad. Nat. Sci. Phila. VI, 1853, p. 413. “Hab, In Georgia.”—Barrb, Rep. U. S. and Mex. Bound. Surv. II, Mamm. 1859, p. 43 (in text). Reithrodontomys humilis RHOADS, Proc. Acad. Nat. Sci. Phila. 1894, p. 161. (Tarpon Springs, Fla.)—CHAPMAN, Bull. Am. Mus. Nat. Hist. VI, 1894, p. 338. (Enterprise, Fla.) Adult.—Above ruddy fuscous brown, usually a little darker along the median line of the back, lighter and more fulvous on the sides, forming an ill-defined fulvous border at the junction of the dorsal and ventral areas. Below dingy gray, usually with a tinge of fulvous, particularly over the pectoral region, where there is a tendency to an ill-defined chest-mark. (The plumbeous basal portion of the fur shows more or less through the grayish tips of the hairs, which, as already said, often present a distinct wash of brownish fulvous.) Feet whitish ; ears more or less dusky ; tail more or less distinctly bicolor, dusky above, gray- ish white below, thinly haired. Immature.—Darker and more plumbeous above, with little or none of the brownish wash of the adults; below plumbeous, washed with whitish gray. Very young specimens are much darker and more plumbeous than those nearly full grown. 1895. | Allen, Species of the Genus Reithrodontomys. | Measurements.—TYail slightly less than half (about 48 per cent.) of the total length. Length, 120; tail vertebrae, 56; hind foot, 15.5; ear, 9.5. (For measurements of additional specimens see Table I, p. 141.) Geographic Distribution.—Coast district of South Carolina and Georgia, and southward into Florida (Enterprise, Chapman ; Tarpon Springs, Rhoads), SPECIMENS EXAMINED. No. of Locality. Date. | Collector. specimens. 5 | Baceboro, Gay «.|) April 12—14.....)..- V. Bailey & R. J. Thomp- | son.! I Bnfenpnseye nua LEDs 27). resis. ost « | C. L. Brownell.’ 13 Raleigh, N. C..; Nov. 11—Dec. 15. ..| H. H. and C. S. Brimley.” 20 me Sel Ove 7) all. 20.5. - oa ' 28 a =~ .2| Dees Beb:, March; 65. April and July. .. a a 1 Received from U.S. Dept. Agr. * Collection Am. Mus. Nat. Hist. 3 Received from Dr. C. Hart Merriam. There is considerable seasonal variation in color and condition of pelage, most adult November and December examples being in short, thin pelage and of a lighter, more chestnut-brownish hue than February and March examples, as shown by the large series from Raleigh, covering the period from Noy. 5 to April 7. Half- grown young differ markedly from the adults in being nearly uniform dark plumbeous. Messrs. H. H. and C. S. Brimley, in answer to my inquiries as to the distribution of this species, have kindly written me as fol- lows: “ The only places in North Carolina from which we have seen specimens are Raleigh and Wolke, in Bertie County, on Albemarle Sound.” They further state that Mr. C. S. Brimley collected in 1890 at Greensboro, Alabama, and at Bay St. Louis, Hancock Co., Miss., without meeting with this species. “At Raleigh,” they add, “it inhabits the upland fields, and also the edges of marshes, but is never found in woods nor in wet mead- ows, where Arvicola riparitus abounds. ‘The few nests that have been found were in damp places in tussocks of grass or rushes. At Raleigh it is one of our commonest mice.” While in general the description of AZus humulis Aud. & Bach. applies satisfactorily to the species of Reithrodontomys occurring near the coast in South Carolina and Georgia, it is singular and 118 Bulletin American Museum of Natural History. |Vol. V1, noteworthy that these authors failed to mention the grooved incisors in any of the three descriptions given by them of this species ; especially when they so particularly refer to the charac- ter of the molars, which they compare with those of A/us and Arvicola, remarking (Quad. N. Am., II, p. 106) “that there are angular ridges on the enamel by which it [this species] approaches the genus Arvico/a ; it is in fact an intermediate species, but in the aggregate of its characteristics perhaps approaches nearest to Mus, where for the present we have concluded to leave it.” They also state that they believe “this animal can be traced as far to the northeast as the State of New York, several having been procured in traps on the farms in the vicinity of the city.” These statements, taken with the fact that W7us humud/is, in their ‘Descriptions of New Species of Quadrupeds inhabiting North America’ (Journ. Acad. Nat. Sci. Phila., VIII, pp. 280-323), is separated from their Jus carolinensis and Mus lecontit by the intervention of Mus aureolus and Mus michiganensis, and the fur- ther fact that grooved incisors are particularly mentioned in the case of JZ. carolinensis and M. lecontit, seem to throw doubt upon the tenability of the name Aumulis for any species of Reithro- dontomys. It is further to be noted that Le Conte, in his remarks upon North American Muridez (Proc. Acad. Nat. Sci. Phila., 1853, p. 410), says that “the Mus Lecontei of Bachman....is a Reithrodon, and neither a Mus nor a Hesperomys.’’ In the same paper (p. 413) he refers both Mus humulis and Mus carolinensis, with which he says he has long been ‘well acquainted,” to Flesperomys, and gives under Reithrodon only R. Lecontit. It is suggestive also that Baird in 1859 (Mex. Bound. Surv., l. c.) compared his 2. megalotis with R. lecontit, and made no mention of &. humulis, the inference from which is obvious, as he had previously considered /econtiz to be a pure synonym of Aumulis. The only objection to referring JZus carolinensis Aud. & Bach. to “ Hesperomys,” as was done by Le Conte, is the statement that “the upper incisors are slightly grooved;” in JZ. /econtii they are said by the same authors to be “deeply grooved.” The distinction here made is noteworthy, especially as the proportions and color of AZus carolinensis accord well with those of a young 1895. | Allen, Species of the Genus Reithrodontomys. [1g Peromyscus (=Sitomys), and do not coincide with any known form of Retthrodontomys from “‘the maritime districts of South Carolina.” Of the pertinancy here of AZus /econtit there is no question. Reithrodontomys merriami,' sp. nov. MERRIAM’S HARVEST MOUSE. Similar in general features to 7. /econtiz, but distinctly smaller, with slightly shorter tail, and much darker coloration. Adult.—Above yellowish gray brown, darker along the middle of the dorsal area, forming a broad blackish band from the shoulders posteriorly ; sides more yellowish gray, with a faint pale buffy lateral line. Below whitish gray, with often a faint buffy wash, most pronounced on the breast. Ears small, uniform blackish ; feet dingy gray; tail very indistinctly bicolor, blackish above, dusky gray below, thinly haired, the annulations often distinctly visible. Measurements.—T ype No. #7323, Nat. Mus , 4 ad.,Austin Bayou, near Alvin, Texas, March 15, 1892; Wm. Lloyd. Length, 112; tail vertebrae, 55; hind foot, 16.55 ear, 9. — Nine specimens (U. S. Dept. Agr.) from Austin Bayou, near Alvin, Brazo- ria Co., Texas, measure as follows: Length, 112 (106-128); tail vertebra, 52 (45-60); hind foot, 16.2 (15.5-17); ear, 8.5 (8-9); ratio of tail vertebrz to total length, 46.4 (44-49). Geog. Dist.—Coast district of southwestern Louisiana to Brazoria Co., Texas. Material Examined.—Austin Bayou, near Alvin, Texas, March 13-17, Wm. Lloyd (U. S. Dept. Agr.), 9 specimens; Lafayette, La., May 22, R. J. Thomp- son (U. S. Dept. Agr.), I specimen. Total, Io specimens. I also refer to this species a specimen (alcoholic) recorded by Coues (Mon. N. Am. Roden., p. 126, Table xxxiii, third line from bottom), under Ochetodon humilis, from Calcasieu Pass, La., of which he gives the following measure- ments (here reduced to mm.): Length, 112; tail vertebree, 56 ; hind foot, 15.2; ear, 9.4. This (if properly referred) forms the first reference to the occurrence of this species in Louisiana. Fortunately the Riceboro specimens of &. /econtii are strictly comparable, as regards season of capture, with the series from Austin Bayou. ‘The differences in coloration are striking ; there 1 Named for Dr. C. Hart Merriam, Chief of the Division of Ornithology and Mammalogy, U.S. Department of Agriculture. 120 Bulletin American Museum of Natural History. |Vol. V\I, is also a quite noteworthy difference in size, and in the ratio of tail vertebree to total length. ‘The pelage is softer and fuller, and the tail more scantily haired. It needs comparison with no other species thus far known. Reithrodontomys dychei,’ sp. nov. DycHE’s Harvest MOUSE. Ochetodon humilis Cours, Proc. Acad. Nat. Sci. Phila. 1874, p. 185 ; Mon. N. Am. Roden. 1877, p. 123 (Kansas, Missouri, Iowa and Nebraska specimens only). Somewhat resembling A. megalotis, but darker, slightly smaller, and with smaller and more distinctly spotted ears. Adult,—Above mouse gray, rather conspicuously lined with black, lighter and more fulvous on the sides, with an indistinct pale fulvous lateral line extending from the cheeks to the base of the tail ; sides of the nose, lower edge of the cheeks, throat and whole lower parts whitish, the hairs being plumbeous at base and broadly tipped with white, without any tinge of fulvous on the breast or elsewhere on the ventral area. Ears of medium size for the genus, well rounded apically, moderately well clothed with short grayish-brown hairs on both surfaces. A more or less distinct dusky spot on the outer edge of the ear near the base, and another at the base of the ear internally, both often obsolete in old specimens. Usually a quite noticeable tuft of yellowish-brown hairs in front of the anterior base of ears. Tail well haired, distinctly bicolor, the upper third dusky and the rest whitish or grayish white. Upper surface of all the feet whitish. Young.—Darker and more mixed with blackish above, with the fulvous lateral line (in middle-aged specimens) more uniformly present and stronger than in adults. The dusky ear spots are more distinct, usually forming rather conspicuous markings. Measurements.—Type, No. '2A2;7, Am. Mus., 9ad., Lawrence, Kans., Jan. 12, 1894; Prof. L. L. Dyche. Length, 133; tail vertebrae, 52; hind foot, 15.5; ear, 10. Twenty-four adults from Lawrence, Kans., measure: Length, 130 (119- 149) ; tail vertebrae, 60 (51-70) ; hind foot, 16.8 (15.2-18.8); ear (from skins), 10 (g.5-10.5) ; ratio of tail vertebrze to total length, 46 (44-49). Five specimens from Neosho Falls, Kans., measure :? Length, 118 (102- 123) ; tail vertebrae, 53 (43-58.4) ; hind foot, 16 (14.5-16.3). The Onaga and Trego series are unfortunately not accompanied by measure- ments. 1 Named for Prof. L. L. Dyche, University of Kansas, Lawrence, Kansas. 2 Measurements from Coues (Mon. N. Am. Roden., p. 126), reduced to mm. 1895. | Allen, Species of the Genus Reithrodontomys. I21 Geog. Dist.—Eastern Kansas, from about the middle of the State eastward to St. Louis, Mo., and from Neosho County north to eastern Nebraska and southwestern [owa. SPECIMENS EXAMINED. we & ov 6.6 Locality. Date. Collector. Whence received. “3 37 |Lawrence, Kan...|Dec. 28-Jan.18, Mar. 22, Apr. Bis CTE OE 'Prof. L. L. Dyche ..!Prof.L.L. Dyche. palNeosho Falls, Kan.|.0..... 52... Col™sNeS: (Goss: .-. (U.S. Nat. Mus. 9 |Onaga, Pottawato-|Oct. 6, Nov. 17- mie Go Kan 2.-| 25, Mec:2; Feb: rr, Apr. 14. .|/F- F. Crevacceur...|/U. S. Dept: Agr. BalOtae LOUIS. MO. 2. il\cjercete ate. on ce ..|Dr. Geo. Engelmann) U.S. Nat. Mus. 2 |London, Lancaster POs NED... -.- Aprilotrrss .. Geo. A. Coleman...}U. 5S. Dept. Agr. 51 This species is much larger than either A. merviami or R. leconttt, ‘Vhe pelage is very long, soft and full. In coloration Z. dychet differs from &. merriami in being much paler, with much less black over the dorsal area, more yellowish gray sides, and clear white underparts, with a much more sharply bicolored tail, and spotted ears; from A&. Zecontiz in much lighter and wholly differ- ent coloration, in much fuller, softer pelage, in its heavily-clothed tail and distinctly spotted ears. Reithrodontomys dychei is based primarily on the large series from Lawrence, Kansas, received from Prof. L. L. Dyche. I refer also to this species the two specimens (one of them is before me) from St. Louis, Missouri, doubtfully assigned by both Baird and Coues to 2. humilis (=Jecontii); also the five specimens (two are before me) recorded by Coues under the same name from Neosho Falls, Kans.; and the single specimens from Burlington Kans., and Buchanan Co., Iowa, similarly recorded by the same author. The Onaga specimens are very dark and very small; some of them are obviously quite young, and all are apparently more or less immature, which probably explains their small size and dark coloration. 122 Bulletin American Museum of Natural History. |Vol. VII, Reithrodontomys dychei nebrascensis, subsp. nov. NEBRASKA HARVEST MOUSE. Differs from 2. dychet in slightly larger size, relatively larger ears, and more strongly fulvous coloration. Adult,—Above yellowish brown finely lined with blackish tipped hairs, par- ticularly over the median area ; the fulvous brown tint is strongest on the sides and posterior half of the dorsum; beneath white. Ears indistinctly spotted. Young.—Above pale buffy gray, faintly lined with dusky hairs ; below white. Ears distinctly spotted. Much lighter colored above than the young of 2. dychet at corresponding ages. s634, Nat. Mus., 4 ad., Kennedy, Nebr., vength, 130; tail vertebrae, 61 ; hind foot, y Uren iT A Measurements.—TVype, No. § — April 19, 1890; Vernon Bailey. 18; ear (from skin), IT. Thirteen specimens from Kennedy, Nebr., measure : Length, 135 (126-139); tail vertebrae, 63.6 (59-68); hind foot, 17.6 (17-18); ear (from skins), 11 ; ratio of tail vertebrze to total length, 45.6 (42-47). . Four adult specimens from Cafion City, Col., measure: Length, 141 (128- 153); tail vertebra, 64 (58-68); hind foot, 16.3 (16-17) ; ratio of tail vertebrae to total length, 45.4. Geog. Dist.—Western border of the Plains, from Fremont Co., Col., north to Custer, Mont., and east to central and northeastern Nebraska. SPECIMENS EXAMINED. 35 38 Locality. Date. Collector. Zo a 14 | Kennedy, Cherry Co., Nebr..| Apr. 19-25.... ...| Vernon Bailey.! 1 Ghency Gol Nebra noc eiol| AMINES Oo 6 ooo 08 A. B. Baker.! 2| Alliance, Boxbutte Co., Nebr.| July 13........... Dr A; Keebishers ie | leauge IekKOlhe (COs. ING bites BP MIEN, occ cepoormes V. Bailey.! 3 | Callaway, Custer Co., Nebr..| Sept. 13, 14..... Geo. A. Coleman. ! 3) Columbus; Platte Cor Nebr.) Ane. 2)7—20)o a ae a Se i weameyn buttaloy@or Nebr) Sept. One) elec ee os 2 | Pendennis, Lane Co.,-Kans..| May 8............ W. W. Granger.’ af |) Aiceyers) (Coys IM cn Se ee IDO OUE Se ob saint 5 - ae tee Wien tenis Dec. 9-29° fo" 3 Bs Bag VE NE Sears Jans 24 ic oiad ae A. B. Baker.! Ral Ganone Citye: Glee iertrie cyte Oct. 2=6.iss sree J. A. Loring.! 1 | Loveland, Larimer Co., Col..| Oct. 22............| C. P. Streator.! 8 ys o ‘* + ..| Apr. 2-14, Sept. 152) We Gp Smuthes Py elle sh ourche iver js. Oalkew.)\s |e 2) ilar ialatr tenets Vernon Bailey.' in|] WSerovilitoya YS IDEs Ga Ano one Dec: cen Eee G. S. Agersborg.* my |. Guster Montes. acter vests June’ t04), ces ee J. A. Loring.! 43 1 Received from U.S. Dept. Agr. 3 Received from Gerrit S. Miller, Jr. 2 Collection Am, Mus, Nat. Hist. 4 es ‘“* Dr. C. Hart Merriam. , 4 1895. | Allen, Species of the Genus Reithrodontomys. 123 This subspecies differs from 2. dychei in its slightly larger size, slightly larger ears, and very much stronger suffusion of fulvous. In coloration it is parallel to the phase of Peromyscus found over the same region, and known as P. americanus nebrascensis, as compared with other conspecific forms of the latter group. 2. dychet and R. dychet nebrascensis undoubtedly intergrade, from the nature of their distribution, over the central portions of Kansas and eastern central Nebraska. This form is based primarily on the series from Kennedy, Nebraska—the only series of which measurements taken from the fresh specimens are available. The Loveland series, however, is quite similar in coloration and apparently in size. One specimen of the latter (No. 495, Coll. G. S. Miller, Jr., April 8) is remarka- ble for its pallor, having an exceedingly bleached appearance, and especially for the absence of the usual dusky stripe along the upper surface of the tail. Another (No. 65,667, Dept. Agr.) from Belle Fourche River, S. Dak., June 2, is remarkable also for its pale gray tint, through, apparently, the fading out of the fulvous tinge so prominent in early spring specimens from other localities. August and September specimens are darker and less fulvous than spring examples. In coloration, hairiness of the tail, and in general features, this subspecies bears a close resemblance to the R. megalotis group, Kennedy specimens finding their exact counterpart in coloration in specimens from Inyo Co., California, and southern Nevada, while specimens of the grayer style are almost indistinguishable in coloration from the phase of the &. megalotis group represented in the San Juan region of New Mexico and Utah. Reithrodontomys montanus (Zaid). MOUNTAIN HARVEST MOUSE. Reithrodon montanus BAIRD, Proc. Acad. Nat. Sci. Phila. 1855, p. 335. ‘‘Col- lected in the vicinity of the Rocky Mountains, lat. 38°;” Bairp, Mam. o»” N Am. 1857, p. 449. ‘* Rocky Mountains, 39°. ? Ochetodon montanus (sp. proband.) CouEs, Mon. N. Am. Roden. 1877, p. 130. (From Baird.) Reithrodontomys montanus ALLEN, Bull. Am. Mus. Nat. Hist. V, 1893. p. 80. (Based on an examination of the type of the species.) “Tail very little less than head and body, which barely exceeds two inches. Hind foot, .50. Ears small, the membrane thickened, and with long coarse 124 Bulletin American Museum of Natural History. |Vol. VU, hairs. Above, brown and pale yellowish gray, much lighter than mouse-color. Outside of ears and flanks, pale yellowish brown, without any rufous. Beneath, —Baird, N. Am. Mam., p. 449. ” duli whitish. ATeasurements.—‘* Nose to occiput, 10 lines; nose to root of tail, 2 in. 2 lines [=51 mm.]; tail, from root to end of hairs, 2 in. [=50.8 mm.]; ears, height posteriorly, 314 lines [=7.4 mm.]; ears, height internally, above notch, 4 lines [=10.2 mm.]; ....hind foot, from heel to end of claws, 6 lines [=12.7 mm.] ; skull, length, 914 lines [=44.5 mm.]; ....”—Batrd, |. c., p. 450. Geog. Dist.—Known only from the type, taken in lat. 38 to 39, in the Rocky Mountains, probably near the upper end of the San Luis Valley in Colorado. Although the original type of the species is before me, I have preferred, owing to its present deteriorated condition, to copy Baird’s excellent description rather than to give a new one. There is nothing, in fact, to be said in amplification of what Baird wrote, the type still remaining unique. There are no specimens in the material before me from any point nearer the type locality than Cafion City, some fifty miles to the eastward, and in a quite different region. ‘The type specimen, as said by Baird, “appears quite adult’; in fact, the teeth are considerably worn, and there are other indications of full maturity. Yet the specimen is not larger than quite immature (one-half to two-thirds grown) exam- ples of 2. megalotis or R. dychet. ‘The very small ears, with the membrane thickened and covered with rather coarse yellowish hairs, the small size of the auditory buile and their rather oblique position, and the rather peculiar enamel pattern of the molariform teeth, are features not seen in any other example of the genus I have examined. In external characters—as the relative length and hairiness of the tail, and in coloration—there is little besides the small thick- ened ears to distinguish it from immature examples from north- ern New Mexico, Colorado or Kansas. The type of Reithrodon montanus is No. 13 of the specimens taken by Mr. Kreutzfeldt on Capt. E. G. Beckwith’s Expedition from Wesport, Mo., to the Pacific Coast in 1853-4. Only a few speci- mens of mammals and birds appear to have been collected on this particular expedition, as on careful collation of Baird’s famous Vols. VIII and IX of the Pacific R.R. Explorations and Surveys, 1895. | Allen, Species of the Genus Reithrodontomys. 125 I find only about a dozen localities mentioned from which speci- mens are credited to Beckwith’s Expedition. ‘The locality of the type in question appears not to have been accurately known even to Professor Baird, who records it as “‘vicinity of the Rocky Moun- tains, lat. 38°,” in his first description of the species, and later as “Rocky Mountains, 39°.” Specimens Nos. 15-18 of Beckwith’s Collection are given as from Sewatch Pass, and Nos. 14 and 20-22 as from Coochetopa Pass. The series begins with No. 1, taken at Bent’s Fort, on the Arkansas River; No. 3 was from the Greenhorn Mountains ; No. 5 from Sangre de Christo Pass, and Nos. 7 and 11 from near Fort Massachusetts. From the itinerary of the expedition (P. R. R. Expl. and Surv., II, pp. 1-128, and particularly pp. 116 and 120-122) it is evident that No. 13, the type of Reithrodon montanus, was taken about August 29 or 30 in the upper part of the San Luis Valley. Until this region has been thoroughly explored for ‘topotypes’ of A. montanus, it would be obviously improper to reject this species as unidentifiable or to give the name precedence over &. mega/otis for the form here recognized under that name. Reithrodontomys megalotis (Auirz). BiG-EARED HARVEST MOUSE. Reithrodon megalotis BAIRD, Mam. N. Am. 1857, p. 451; Rep. U.S. and Mex. Bound. Surv. II, Mamm. 1859, p. 43. Between Janos, Sonora and San Luis Springs, New Mexico. Reithrodontomys megalotis ALLEN, Bull. Am. Mus. Nat. Hist. V, 1893, p. 79 (San Juan region of New Mexico and Utah); ALLEN, ibid. VI, 1894, p. 320 (Fairbank, Arizona). Reithrodontomys aztecus ALLEN, ibid. V, p. 79. La Plata, New Mexico. Ochetodon humilis Cours, Proc. Acad. Nat. Sci. Phila. 1874, p. 185; Mon. N. Am. Roden. 1877, p. 123. (Only the references to 2. megalotis Baird. )— Aston, Biol. Centr.-Am. Mamm. 1880, p. 151. (The Sonoran references only.) ‘“ Largest of North American species. Head and body from 2.50 to 3.00 inches [—63.5 to 76 mm.|; tail about two-tenths shorter. Hind foot near .70 {[=17.8]. Ears large, moderately clothed with hair. Above mouse-gray, lined with darker, and tinged with rusty ; on the rump and sides a fulvous wash. Beneath soiled yellowish white.” —Aazvd, Mam. N. Am., p. 451. The above is an excellent description of average adults. Immature speci- mens are grayer with less of the fulvous wash, and with indistinct blackish ear spots, asin 2. dychei, Occasionally the fulvous on the back in adults shades 126 Bulletin American Museum of Natural History. [|Vol. VII, on to areddish tinge. The upper surface of the tail is distinctly darker than the sides and lower surface, well clothed with short hairs, wholly concealing the annuli, except in worn specimens. Feet soiled whitish. Measurements.—(See Table II, p. 141.) In compiling the table obviously immature specimens were excluded, although a number of * young adults’ are embraced in several of the series, as in those from Aztec and Provo, thus tending to lower the general average. Geog. Dist.—Northeastern Sonora northward through western New Mexico and eastern Arizona to Northern Utah. ‘The localities given in the following table indicate more in detail the known distribution of the species. SPECIMENS EXAMINED. noe Locality. Date. Collector. mens. t | NearSanduis Springs; Sonora: |soen sou eect Dr. C.B. Kennerly.! I Orie Ua chucae Atiz mtr Sed |e ere sere tees Dre 2. BY Wilcox 5. | RainbankayAtiz a5 eee see “March 2-I4.. Price and Condit.® 1 ieity AGG Nin soa caosene JENN Pio). SB aae ee Vernon Bailey * T "| silver City NEG Mlexeri: n= a DeEC barrier C.-P. Streatome I WasiViemasta Nie Neem. nee April 6): cn. shy silo sek ee z 2 |) dA. ING MIG a5 pop eoseec March 19, 20...| C. P. Rowley.? 33 Se re Eee oe DEC 5-0 seca J. A. Loring.? 7 ik live, Plata ING Mier corstcie net Mch. 30-Apr.11.| C. P. Rowley.” 34 ss a Serr iit Wee. TO-T2s. 5s J. A. Loring. I Riverview. Uitahe-e-m— users April 25a raster 'C. P. Rowley.? r | Blatt City, Utah. 325.2055: May 8822s [sae rh 4| a a ees eRe eh Nov 8.705 seer J. A. Loring.4 4 Noland’s Ranch, S. W. Utah..| Nov. 23..-..... < ss 3 Fairfield, Utah..... ..-.-..| June 24, 25....] Vernon Bailey.4 I Manti, Utah. . Peat ae Deco 0s. See. alla eee eee 2 1 | Camp Floyd, Wiehe, ca. etude ae Cas vice arthy. |i slin Cowes Wiel. cope - Se PATNI eerie pores Vernon Bailey.” 15 Provo; Uitahiea. fate- 6 = Nov. 11- Dec. its se oe 2 Latvon, Witkin see on auoe b\| OCU C2 Rees ae fe ss Be Osdenn Utama meres ttey- tars Oct. 1=3 5 =e a a 126 1 Received from U.S. Nat. Mus. 3 , Receiv ed from Dr. C. Hart Merriam. 2 Collection Am. Mus. Nat. Hist. * U.S. Dept. Agr. R. megalotis was described from two specimens (Nos. 1039, skin, and 1040, alcoholic, U. S. Nat. Mus.) taken by Dr. C. B. R. Kennerly, between Janos and San Luis Spring, Sonora, near the boundary line of southwestern New Mexico, the former of which is properly to be considered as the type. Of these specimens only the skull of No. 1039 appears to be extant. ‘This specimen, ee ee Ee ee eeimmetind 1895. | Allen, Species of the Genus Retthrodontomys. 127 through the kindness of Mr. True, Curator of Mammals in the U.S. National Museum, I have the opportunity to reéxamine’ in the present connection. ‘This skull, taken in connection with Baird’s excellent description of the external characters, appears to leave no question of the propriety of applying the name to the species now so well represented by specimens from New Mexico, Arizona and Utah, among which is a small series from Fairbank, Arizona, a point about one hundred miles north- west from the type locality of the species. ‘The Fairbank speci- mens are not appreciably different from large series from the San Juan Valley in northwestern New Mexico and southeastern Utah. Specimens from central and northern Utah are so closely similar that I am unable to specify any differences. In a series of 13 specimens from Winslow, Arizona, the tail averages slightly longer than in any of the series from New Mexico and Utah, and on this account has been referred to the next form rather than here. Reithrodontomys megalotis deserti, subsp. nov. DESERT HARVEST MOUSE. Similar in coloration to R. megalotis, but with a considerably longer tail. Tail 50 to 52 per cent. of the total length, instead of 46 to 48 per cent., as in R. megalotis proper. Type, 333575, U. S. Nat. Mus. (Dept. of Agr. Coll.), 9ad., Oasis Valley, Nye Co., Nevada, March 16, 1891 ; F. Stephens. Measurements.—(See Table III, p. 142.) Geog. Dist.—Southern Nevada and Inyo Co., California. Specimens Examined.—(See next page.) In coloration, general size, size of the ears, hairiness of the tail, and in other external features, there is very little difference between examples from the Death Valley region of California and adjoining portions of Nevada and specimens from northern Utah and thence southward to western New Mexico and eastern Ari- zona. The fulvous suffusion of the dorsal surface is possibly a little stronger and more of a brownish cast than in the Death Valley specimens, but the average difference in this respect is so slight as to be thoroughly masked by the wide range of individual and seasonal variation shown by any of the larger series, and 1 See this Bulletin, V, 1803, p. 70. 128 Bulletin American Museum of Natural History. [Vol. VU, SPECIMENS EXAMINED.’ : speci- Locality. Date. Collector. mens 13.) | Winslow, Amiz. 0.2 446 Manyara ae C. P. Streator. 2 | Grapevine Mts., Esmeralda Cox INGWio ca aGp ec coesognesn bond Mch.22 & June 9} F. Stephens & E. W. Nelson. 26 | Ash Meadows, Nye Co., Nev....| March 4-12....| Fisher, Stephens, Nelson & Palmer 16 | Oasis Valley, Nye Co, Nev......| March 15-18.. -| F. Stephens. L | |sPanaca.leincolny Cory Neva ese) i cay, 20 reper rier Vernon Bailey. 12 | Pahrump Valley, Lincoln Co., Nev.| Feb. 17—Mch. 16,| E. W. Nelson. 4 | Pahranagat Valley, Lincoln Co., INGA ga cdcooue foued cothecs May 24-26..... Vernon Bailey. 5 | Vegas Valley, Lincoln Co, Nev..| March 11-16. .. a 30) || one Pine wiinyoiCowCallaee see Dec. 5-17 and | Bailey, Nelsonand tine 7-0) eer Fisher. 5 Keeler pimyoi Gory Calera ee Dec. 8-I0...... E. W. Nelson. 6 | Owens Valley, Inyo Co., Cal....| June 26- ee 10;| F. Stephens. 1 | Emigrant Springs, Inyo Co., Cal. April 15. I | Twelve Mile Spring, Inyo Co., Calleectacinn darn ener eee ebuer, <1et-te E. W. Nelson. 5 | Grapevine Ranch, Inyo Co., Cal .| April 2-4....... F. Stephens. 20 | Olancha, Owens Lake, Inyo Co., Cali iriganc: tape aero May 16-22. .. os 2 | Ash Creek, Owens Lake, Inyo Co., Calor ce heigs eter: ake eae eee MAP SIO). 6 dla eB oc ES Io | Cartago, Owens Lake, Inyo Co., Calbss 2.5 55 tiie cel seco aes esd rere oe g: | Panamint Valley, Inyo Co., Cal..| Jan. 8-10o...... Bailey and Fisher. 6 | Panamint Mts., Inyo Co., Cal...) April 5— Re 277) 3\ Brae _Nelson. I | Shepherd Cajfion, Inyo Co., Cal..| Jan. 3..... 3 | Resting Springs, Inyo Co., Cal..| Feb. 9-18 .. ..| Fisherand Nelson. 3 Satatova springs) bnyvo Con Galee sheet ony nrre Vernon Bailey. 5 | Death Valley, Inyo Co., Cal.....| Feb. 3 & June 20,| Fisher and Bailey. L )|eArousevitss myo Cor Galhar. April 25........| F. Stephens. I Furnace Creek: Inyo)Co:, Calé=— =| Aprallio... 2. bd I Bishop Creek, Inyo Co., Cal. ...} August 9....... 189 1 All received from U.S. Dept. Agr. therefore is scarcely serviceable as a diagnostic feature. On the other hand, the difference in the relative length of the tail seems too important to be ignored, amounting to at least 4 per cent. Thus in 2. megalotis the tail is decidedly less than half the total length, while in 2. megalotis deserti it equals or exceeds one-half the total length. In a series of about 100 examples of 2. mega/otis proper it is exceedingly rare to find one in which the tail verte- bree equal one-half the total length, while in a still larger PR neg mn gt ne oe LE TO EP ee ——= 1895.] Allen, Species of the Genus Reith: odontomys. 129 ». series of deserti specimens rarely occur in which the tail length does not equal or slightly exceed half the total length. In the average there is a difference of 4 per cent. in the relative length of the tail in the two forms, with only an exceptional overlapping by individual extremes. These differences are well shown in the tables of averages and extremes of the two forms (see pp. 141, 142). A small series of five specimens from the Panamint Mountains, Inyo Co., California, seems to offer an exception to the general rule obtaining in the series from neighboring localities, in this series the tail dropping down to the length proper to 2. megalotis. Whethei a larger series might not alter this ratio, or whether the series indicates a local short tailed form within the range of the long tailed style cannot at present be determined. The series from Winslow, central Arizona, agrees so well in tail-length with the desert group that it is provisionally referred heie, although on geographical grounds it would seem more natu- rally referable to true megalotis. ‘The coloration is also slightly different from that shown in series from other points, so that possibly the Winslow series may indicate the presence of a slightly differentiated local phase in central Arizona. There is quite a wide range of variation in the coloration of the upper parts in specimens from the same locality, strictly compara- ble as to sex, age and season, specimens varying from pale grayish brown, washed with fulvous to much darker grayish brown washed with dark cinnamon. Reithrodontomys longicauda (Aazrz). SONOMA HARVEST MOUSE. Reithrodon longicauda BAIRD, Mam. N. Am. 1857, p. 451. Petaluma, Sonoma Co., California. Ochetodon longicauda COvuES, Proc. Acad. Nat. Sci. Phila. 1874, p. 186; Mon. N. Am. Roden. 1877, p. 126. Adult.—Above yellowish brown, heavily lined with black, the profusion of the intermixed black hairs usually forming a distinctly blackish area along the median line of the back; sides paler, less blackish and more yeliowish, with a rather broad fulvous lateral line, extending from the cheeks to the rump, vary- ing in intensity and distinctness in different individuals. Below dingy grayish white, often with a tinge of yellow, and sometimes with a more or less distinct fulvous patch on the breast. Ears dusky, thinly clothed with yellowish brown [ May, 7895.] 9 130 Bulletin American Museum of Natural History. |Vol. VII, hairs, with, as in nearly all the forms of the genus, a tuft of rusty brown hairs at the anterior base. Feet soiled white ; tail rather sharply bicolor, dusky above, grayish white below, covered with short hairs, only partly concealing the annuli. Young.—Above nearly uniform mouse color ; below plumbeous, washed with gray. Ear with an elongated dusky spot near the base of the outer edge, which is usually obsolete in adults. Young adults are duller and more uniformly colored above than adults, and generally lack the blackish median area along the back. Summer and early autumn specimens are usually paler, with less black, than winter and early spring specimens. Measurements.—Average and extremes of 6 specimens from Glen Ellen, Sonoma Co., Cal.: Length, 136.5 (129-144) ; tail vertebree, 72 (68-79); hind foot, 17 (16-18) ; ear (from skins), 11.2 (10.5-12); ratio of tail vertebrz to total length, 52.2 (50-54.8). This is about an average series, from near the type locality. Measurements of other series are given in Table IV (p. 142). The tail varies from 50 to 56 per cent. of the total length, averaging about 53 per cent. SPECIMENS EXAMINED. 35 6 Locality. Date. Collector. Ao ay As kchamas dichamealCoyn@alearner IDaeh BON Silo oone C. P. Streator. Gn eesvalleGolusal Conn Callie (umes — 8 er F. Stephens. ' 3 | Lower Lake, Lake Co., Cal..... Aprile 2325 eee Sy ro} Glen Bllen; Sonoma 'Co:, Gall. -=) Jan, 10-29-27: 2.2 C. P. Streator.! 3) | Hairtield ssolanolC€o.) Cale wee Neb A =O tb teeiee a - 2;| Novato; Marimi Cor, Galen fain S828 sapere eeeenaer: es e 5 |) Nicasio, ManmiGomyCallmine oc: March6-9....... C. A. Allens: 5 ne e Rp erty wae Rebs 7523s C. P. Streator.! I | Jackson, Amador Co., Cal...... Manchisa\eaene es 1 | Martinez, Contra Costa Co., Cal.| April 13......... F. Stephens.! 20 | Walnut Creek, ContraCostaCo.,‘‘ .| Feb. 14-20......| C. P. Streator.! Lo) |) racy. san) JoaquiniCo:, (Cal (||, anos TOs ire J. E. McLellan.! 3 | Berkeley; Alameda Go:, \Cal .-...| Reb: 27-29 0. 241"@> Ra wtreaton: T4 | San Mateo, San Mateo Co., Cal..| Feb. 5-g........ J. E. McLellan.! 11 | Portola, Santa Cruz Co., Cal....| March 20-28.....| R. L. Wilbur.* 45 | La Honda, x ‘** ....| Dec. 24-Jan. 1...} Price & Wilbur.® 16 | Monterey, Monterey Co., Cal ...| Sept. 30-Oct. 2..| Vernon Bailey.' In Rsane | acinto,mWonterey Gor Gal.) © chrono F. Stephens. 2 \ebaciticl Groves si. =. Bi Oet. (Guat wi. eee J. H. Oliver) I | Posts, Ae : - oe arehenien ee sere J. E. McLellan.' 7 \ Sut eS ue - al Marchi4=03t cei oe ve 5 | Jamesburg, me Hi ve |ovlarch! 18—2 eee ee * Lt fie ay ele aa —— ! Received from U.S. Dept. Agr. 2 Collection Am. Mus. Nat. Hist. 3 Received from W. W. Price. 1895. | Allen, Species of the Genus Reithrodontomys. I31 Geog. Dist.—California, west of the Sierra Nevada, from the coast region of Monterey County north to Mendocino County, and in the interior from San Joaquin County north to Tehama County. Probably further south, irregularly, in the Coast and San Bernardino ranges of mountains. Specimens Examined, 175, as shown by the foregoing table. Fourteen of the 15 specimens on which Baird based his 2. longicauda were from Petaluma, and the other from San Fran- cisco, California. As shown above, the region of the type locality is well represented in the present material. Baird’s measurements, based on alcoholics, fall much under the average, taken from fresh specimens, due probably to the presence of a rather large proportion of more or less immature examples. The table of measurements (see p. 142) apparently indicates more or less variation in size with locality, but this is more appar- ent than real, since in the series giving large measurements all of the specimens are practically adult, while those giving smaller averages contain examples that are not fully grown, although as a rule, obviously immature specimens were thrown out in making up the table. Reithrodontomys longicauda pallidus (2/oads). SAN DIEGO HARVEST MOUSE. Reithrodontomys pallidus RHOADS, Am. Nat. XXVII, Sept. 1893, p. 835. Santa Ysabel, San Jacinto Mts., San Diego Co. Cal. (Type in Am. Mus. Nat. Hist.) Paler and slightly larger than 2. Jongicauda. Adult.—Above grayer, less fulvous, and less varied with blackish than 2. longicauda. Averages about 8 to 10 mm. longer, with slightly larger ears, as shown by the measurements (Table V), in comparison with the table for R. longicauda (Table IV). Measurements.—(See Table V, p. 143.) Geog. Dist.—Southern California and northern Lower California, from Monterey County on the Coast and Merced County in the interior southward. Appears to develop slightly differentiated local phases in some of the southern mountain ranges. Specimens Examined.—(See next page.) Respecting Mr. Rhoads’s &. pallidus, | find myself greatly em- barrassed as to which of three courses to pursue in the matter, 132 Bulletin American Museum of Natural History. (Vol. VU, SPECIMENS EXAMINED. No. of speci- Locality. Date Collector. mens 7 : = 2 | Boulder Creek, Monterey Co.,Cal.| Oct. 15.......... Vernon Bailey.! i |ebear Valleys San BenitoiGor) Cal s\timer2 2 eer J. E. McLellan.! I oss Banos Merced! Go. Galina mi) cnens eee : = 30) \sHresnos resno Co @al sea March 3-6....... C. P. Streator. ! 1 uhneeshiverss) dilate Go Gallager ial liya2 ee etter nr TS Palmers | Wemoore, Keamnosi Cons Galera Feb. 27 J. E. McLellan? © |) KermiRiver, KerniG€oiCalkeeee: ethos S33 Poet Vernon Bailey.! 7 | San Emigdio Canon, Kern Co., , Calle etic wegnessucte crereoie sisicer Oct 18, Tor. .| E. W. Nelson.! i |tehachapi isernl Com Calero ei RNS), 5 Ss sh hoe J.E McLellan! Dy | AdobeiStation samen) ase seciricner Oct 3): seen ie E. W. Nelson.! r ORI Fort Rejon, 0) ers aes (hulye aes <2 een or T. 'S. Palmer 5 | San Simeon, San Luis Obispo Cor. Cals ene sen Peo ee INN OP) ao stone de E. W. Nelson.! 5 | San Luis Obispo, San Luis Obispo Goss Calne iias 5 Seeminelepe ~ Nov. 26, 27 Ss Sa ozo; ‘San Luis Obispo Co., Cal. OGt20 7 a. bee or de 2 Morro, INO Va EON > ctr ter ub 2a | SPAS Robles, fa! 4 March 12, 13 F. Stephens. ! 4 | Jolon, oe os March 31-April 2,| J. E. McLellan.! 3 | Santa Maria, te ef Deer 20a visa E. W. Nelson.! 5 Gaviote Pass, Santa Barbara cae Call joa dikeyspeto gavel forte acne = DEC O-T2seralior a 4 | Santa Inyez Mission, Santa Bar- baraGos Call Vy ey acta eat Weet4e savers am coe iG 2ealsanta Barbara, Santa Barbara Cot Cal Saino set Ronee rane. DEC MEQ sce prs F. Stephens.! 2 | Carpentaria, Santa Barbara Com (CPE em aie, sana hale Wa Ss ASDECS TQ) Seine ere E. W. Nelson.! 2 | Los Olivos, Santa Barbara Co., Galliss orocts tate otis Ape eee Miarchy6i> = sci F. Stephens.! 4 | Hueneme, Ventura Co., Cal..... Bie: 2215) crore neta os 3. | Ventura River, Jal a eas Dee, 21-23 ae E. W. Nelson.! 3 | Montalva, oo Veni Re: Reb. e20s meee F. Stephens.! 7 S| Sis Weeruiley, ae pe aS ae Dec. 29-Jan. 4...| E. W. Nelson.! 4 | Burbank, Los Angeles Co., Cal.'.) March 10-12..... C. P. Streator. ! 7 | San Fernando, nf March 18=22..... S 4 | Santa Monica, se Heb; TO—rewseeeyae F. Stephens. ! I Calabasas, re Reb iy2) 7) aceuiee 2 | Las Virginus Creek =| DEED! 22 ecrerarersiewets By 3 Réche Cajfion, San Bernardino Os Call pss) pays eyes hoe se Sept. 22=245.-h 1. \ 2 | San Bernardino Peak,San Bernar- dino; Co Call yt coe es ase Oct. 2. eer J. E. McLellan '! 2 | Elsinore, Riverside (on Cal eect INOW. 22 y= epioseyeyy ‘Oosela, *AqpRo0'T “SNYUDIND snupoixau “oy 410 SNANIOAdS lig dO (SANAULXY GNV SAOVAGAY) SINAWAYNSVAW— JA (GI- 11) 9°11 (Q°Z1-G' OL) 9° IT (g°3I-— 11).2° (ZI- 11) FIT (@°3I- OI) 8°Il (ZI- OI) TI ‘IVY (6-00) 21 (81-91) 41 (SI-91) ¢ (SI-91) AT (41-91) 9°91 (gI-91) 41 100) pul (F9-08) ¢°3a (g9-g¢) 9° #9. (¥9-z4) 9 (ge-09) ¢° 69 (9'@g-1¢) ¢¢ (eg-0¢) Fg sy Suay 1230} 03 [1e1 Jo ONBY (g8-0L) 94 (F8-ZL) 94 (6L-ZL) 94 ie =¢L) i (F8-0L) GL (g8-69) 94 "BAGIWOA [le], (091-G81) SFI (681-GE1) 681 (OG1-GS1) SFI (OGI-FS1) OFL (I9T-SS1L) 11 (@@I-ZSL) FFI *yisugy] [e107 “suoutoads BOON: ave) © eave syne Te oyeOn) o8e1q ueg “oja ‘equinoel eee ey era) osay ueS ‘Jaques v}ULG “or vieqivg vues “eoluo yy, BIULG n - ‘opuvutay ues yep‘ ojvAN}UD A “epNeGeJULG ” ” ” *AU[LIO'T Nee ee eee ee een ee ee eee ee ee ee ee a ‘supyog vpnviisuo] “yo AO SNAWIOAdS $O AO (SANAULXY ANV SAOVUAGAY) SLNAWAXASV ATW — d Vv A Article IV.—ON THE OSTEOLOGY OF AGRIOCHCERUS. 3y J. L. WortTMAN. PLATE I. Although the genus Agriocherus has been known for many years, and has always been abundantly represented in our collec- tions by numerous complete skulls, yet it was not until the past year that we have obtained any information regarding the remainder of its skeletal structure. The first intelligence of the very curious organization of its feet was published by Professor Osborn and myself in the description of a remarkable hind foot’ from the Protoceras layer of the White River beds, obtained by the Museum Expedition in 1892. On account of the large claw- like ungual phalanges, and in the complete absence of teeth, we referred it to the order Ancylopoda, established by Cope, and considered it to represent a distinct subdivision of this group (Artionychia). Professor Scott, upon careful examination of the specimen, shrewdly surmised that the foot probably pertained to a species of Agriocherus. The explorations of the past year have demonstrated the cor- rectness of this surmise, and he has added to our knowledge of the genus by a description of a portion of the fore limb.” He has also, in the same paper, discussed at some length the system- atic position of the genus within the Artiodactyla. Another important addition to our knowledge of the probable ancestral genus has recently been made by Professor Marsh in the descrip- tion of a new form (Hyomeryx breviceps) from the older Uinta beds.* During the past year the expedition from the American Mu- seum into the White River beds, near the same locality where the hind foot was found, was fortunate enough to discover a more or less complete skeleton of Agriocherus latifrons, together with numerous skuils and other important parts of the skeleton of dif- ferent individuals of other species, so that the materials are now 1* Artionyx,a New Genus of Ancylopoda,’ Bull. Amer. Mus. Nat. Hist., Feb., 1893, pp 1-18. * * Notes on the Osteology of Agriochcerus,’ Amer. Philos. Soc , May, 1804, pp. 244-251. 3* Description of Tertiary Artiodactyles,’ Amer. Jour. Sci., Vol. XLVIII, Sept., 1894, pp. 250-274. [| June, 1895.| [145] LO 146 Bulletin American Museum of Natural History. [Vol. VU, at hand to enable me to give a tolerably thorough account of the osteology of one of the species at least. Another considerable addition to the materials that I was fortunate enough to obtain, was found in the Cope Collection, which the Museum has recently acquired, consisting of a complete skull associated with numerous limb bones and vertebre of a single individual, collected by my- self in 1879 in the John Day Basin in Oregon. This specimen has aided me materially in supplying the missing parts in making the restoration. It may be added here that the association of the large claw-like terminal phalanges with the teeth, in at least two of our White River specimens, leaves no room for doubt as to the correctness of the determination that this type of ungual phalanx belongs to Agriocherus. It is the object of the present paper, therefore, to present as complete an account as possible of the osteology of this group, together with a critical review of the species which have been described as belonging to it. Following this I will take up the question of the systematic position of the group. OSTEOLOGY. Skull.—This part of the osteology has been so thoroughly described by Leidy, Cope, and Scott, that little remains to be said concerning it. It may not be amiss, however, to recall] some FY TTT, MU - ens TT Fig. 1. Side view of skull of Agr zocharus major. One-third natural size. 1895. | Wortman, Osteology of Agriocherus. 147 of the more important characters in which it differs from its nearest cotemporary selenodont allies—the Oreodontide—as well as those characters in which it resembles them. The general out- line of the skull is very much like that of the earlier Oreodonts, especially Oreodon culbertsont, with which Leidy compared it in his original description. It is rather elongated and narrow, with moderately elevated, compressed, overhanging occiput. ‘The face is but little bent down on the basicranial axis, and the form and relationship of the facial bones, with the notable exception of the premaxillaries, are practically the same as in Oreodon. The otic bullz are always inflated, they are not filled with cancellous tissue, and the foramina at the base of the skull are similarly disposed as in Oreodon. An apparently constant exception to this latter correspondence, however, is seen in the presence of a moderate sized foramen, generally equal to or slightly larger than the foramen opticum, which opens just in front of the sphenoidal spine, in Agriocherus. It is situated above and a little posterior to the foramen opticum. The office of this foramen, as well as its homology, is difficult to determine, but judging from its size and direction I am inclined to regard it as the foramen rotundum. The principal characters in which the skull of Agriochwrus differs from that of Oreodon may be enumerated as follows : In Agriocherus the premaxillaries are reduced and practically eden- tulous. In our collections there are three skulls of different species, in which these bones are in a good state of preservation, and they show that the premaxillaries were not in contact in the median line ; they are small and project but little in advance of the canines. There is a single, small, shallow alveolus upon either side from which the incisors had apparently been shed early during life. In all the cotemporary Oreodonts, on the other hand, the premaxillaries are well developed; they are in contact in the median line, and always bear their full complement of incisors. Some of the later forms, however, show a marked ten- dency to incisor reduction. In Agriocherus the posterior rim of the orbit is not enclosed by bone, whereas in Oveodon the bony ring of the orbit is com- plete, and there is always a distinct preorbital pit or fossa which is absent in Agréochwrus. In the more primitive Oreodont genus, 148 Bulletin American Museum of Natural History. |Vol. V1, Protoreodon, however, the orbit is open posteriorly as in Ag7zo- cherus, and there is no lachrymal pit. The dentition of Agriocherus presents some striking resem- blances to the true Oreodonts ; in other respects it more nearly approximates /Zyopotamus, while in others still it possesses char- acters peculiarly its own. The most characteristic Oreodont feature is seen in the enlargement of the first inferior premolar into a caniniform tooth, while the true canine is small, incisi- form, and so placed as to form a continuous series with the incisors. ‘The upper canine is large, considerably curved, and has a characteristic D-shaped pattern on cross section, as is seen in all the Oreodonts. The characters in which the dentition of Agriocherus departs from that of Oreodon are especially seen in the presence of a diastema between the canines and premolars in the upper jaw and between the caniniform first premolar and the second premolar in the lower jaw. In Oveodon all the teeth are arranged in a continuous series. The structure of the molars presents many important differ- ences from those of Oreodon ; the crowns are lower, less seleno- dont, the valleys are much more open, and the angles of the superior teeth more rounded off. In Oveodon the external median buttress is compressed from before backwards into a vertical plate, whereas in Agriochwrus it forms a wide loop. If it were not for the absence of the anterior intermediate cusp, the molars of Agriocherus would resemble those of Wyopotamus very closely. The only genus known to me in which the structure of the superior molars is strictly comparable is Aerycopotamus of the Indian Miocene, and it would not indeed be surprising to find, when the osteology of this latter genus is more fully known, that the two are quite closely related. The Vertebre.—There is no single specimen in our collection which contains a complete vertebral column, so that the exact number of vertebrae cannot be made out with certainty. In one, however, in which the limbs are more or less complete, the poste- rior five dorsals, all the lumbars, the sacrum, and nineteen of the caudals are preserved. In this specimen there are six lumbars, and if we allow thirteen as the number of the dorsals, we will 1895. | Wortman, Osteology of Agriocherus. 149 then have the highly characteristic dorso-lumbar formula for all the known Artiodactyla. The at/as presents the same general outline as that seen in the Artiodactyla. The articular cavities for the condyles of the skull are deep and spacious and are overhung by the anterior superior part of the arch. In Oreodon and all the recent genera this part of the arch is interrupted by a wide notch which shortens its fore and aft extent. In Agriochwrus this notch is very narrow, and is continued upwards and backwards as a deep groove which sepa- rates the spine into two low indistinct tubercles. ‘lhe transverse processes are well extended laterally, somewhat broader in front than in Oreodon, and project backwards further behind the facets for the axis. They are perforated by moderate sized foramina for the passage of the vertebral artery, which does not appear to be the case in any specimen of Oreodon which I have examined. Anteriorly, the foramen pig. Top view of atlas for the exit of the suboccipital nerve is %Aszicherus suyotianus. large and conspicuous, while the inferior tubercle is small. The facets for the axis are more transverse, and not so oblique as in Oveodon, resembling more nearly the sheep or deer in this respect. The axis, as described by Scott,’ differs from that of Oreodon. This is especially to be seen in the character of the spine. In Agriocherus it is unusually high and prolonged in front, so as to overhang the odontoid slightly, while behind it is not so produced, reaching no further than the extremity of the posterior zygapophyses. In Oveodon the spine is much lower, but little produced in front, but greatly thickened and extended posteriorly. The odontoid, as already well known, is intermediate between the peg-like form of the pig and the Fig. 3. Side view hollow half-cylinder of the higher forms. In ee co - 2 : Two-fifths natural some of the larger specimens in our collection | size. 1 Beitrage zur Kentniss der Oreodontide, p. 361. It is also stated in the same paper (p. 322) that the atlas of Oreodon has the transverse processes perforated by the vertebral canal. In all the specimens in our collection the transverse processes are imperforate, but the position of the canal is frequently indicated by a pit of variable dimensions. 150 Bulletin American Museum of Natural History. |Vol. VU, the odontoid is almost as highly developed as in any of the living genera. The remaining cervicals are very much like those of Oreodon ; they are provided with prominent hypopophyses and moderately developed neural spines, which increase rapidly in length from before backwards. The dorsals resemble those of Oreodon very closely, so far as can be determined from our somewhat imperfect material of this region. ‘The spine of the first dorsal, however, is much longer than that of the corresponding vertebra of this genus. The posterior six have rather elong- ated, slightly keeled centra, with Fig. 4. Front view Fig.5. Side view nearly flat oval faces. In the of cervical of Ag77- of cervical of Agyz- : ocherusguyotianus. ocherusguyotianus. Ninth, tenth, eleventh and i vofitths natural voniths natural twelfth, the zygapophyses are nearly flat, while those of the thirteenth begin to assume the tongue and groove pattern of the lumbars. The neural spine of the ninth is high and back- wardly directed, that of the tenth being more nearly vertical. From this point backwards the spines have a more forward direction. The transverse pro- cesses begin at the eleventh and Fig. 6. Anterior Fig.7. Sideviewof become more and more promi- view of first dorsal of first dorsal of A grzo- : Agriocherus guyo- cherus guyotianus. nent posteriorly. Metapophy- tianus. Two-fifths Two-fifths natural < atte natural size. size. ses are fairly well indicated on the last two dorsals. The Zumébars are six in number. The second, third and fourth have moderately strong ventral keels, the two last being practi- cally without this structure. The centra increase in size and length from before backwards, the last two being markedly flat- tened vertically ; the central faces exhibit a slight convexity both in front and behind, except that of the last lumbar, where it joins the sacrum, which is nearly flat. ‘The spines are broad and ele- vated, and the metapophyses well developed. ‘The zygapophyses 1895.| Wortman, Osteology of Agriocherus. ISI exhibit a well-marked double tongue and groove (MammMALs.GCornm- LECTED: [The external measurements given in the following list are the collector's measurements taken from the specimens before skinning, unless otherwise stated, and are of course in millimeters. | 1. Dorcelaphus' couesi (Cowes & Yarrow). SONORAN.DEER. Cervus mexicanus BAIRD, Mam. N. Am. 1857, p. 053 (excluding synonyms) ; in part only or not at all the Cervus mexicanus of Gmelin and Jater authors, Baird excepted. Cariacus virginianus var. COUES & YARROW, Wheeler’s Geog. and Geol. Surv. West of tooth Merid. V, 1875, p. 72. “Cariacus virginianus var, Couesi ROTHROCK MSS.” zdid. p. 72, and, by implication, in text, p. 75. A small Deer, evidently the same as Cervus mexicanus Baird, is represented by seven specimens in the Price Collection, six of which were taken in the Santa Cruz Mountains, Sonora, Feb. 12-15 (B. C. Condit), and the other in the Huachuca Mountains, Jan. 28 (Price and Condit). They agree essentially with Professor Baird’s description, based on a female taken at San Luis Springs, Sonora, so far as the description goes. It may be the same also as Lichtenstein’s Cervus mexicanus, described from specimens sent alive to Berlin, in 1825, by Herr Graf, from “ Mexico,” with- out indication of the exact locality at which they were taken. The Cervus mexicanus of Gmelin, however, is a vague composite species, only in part referable to Deer from Mexico, and in all probability has no relation to the little Sonoran Deer described by Baird. The specimens here under consideration are from a point probably not more than fifty miles from the type locality of Baird’s Cervus mexicanus. As this name is clearly untenable in the present connection, the name cowesz, proposed by Roth- rock, may be employed for its designation. Above grayish brown, rather paler than the winter coat of wirgintanus, the hairs being broadly banded with blackish brown subterminally, and tipped with whitish ; darkest along the middle region of the back, paler and slightly yellowish on the sides ; belly white ; axillar region pale buffy ; ears dusky, the 1 According to ne Oldfield hpiiass Diecsithes "Gloues (1842) * panels anid antedates * Cariacus Lesson (1842). C/. Thomas, Ann, and Mag. Nat. Hist., (6) XV, p. 193, Feb., rae id 1895. | Allen on Mammats from Arizona and Mexico. 201 hairs tipped with gray ; a narrow blackish nose band, and a small spot of black- ish on each side of the lower jaw near the end. Tail yellowish brown above (the hairs brownish dusky at base) ; below white ; no black anywhere at the surface of the hairs. Distal half of legs yellowish brown in front, lighter behind. Antlers of the D. wirginianus style, but much smaller, with much shorter tines. The basal point (in four full-grown bucks) varies from one to two inches in length, and the longest point varies from three to four inches. Measurements.—Vhe only external measurements available are: ‘‘ Ear, 140 mm. ; tail, 225.” The skull of an old male (largest of the series) meas- ures as follows: Total length, 246; basilar length, 227; zygomatic breadth, 115 ; lower jaw, length, 188; height at coronoid, 95; height at condyle, 61; distance between base of antlers, 66 ; distance between points of antlers, 283. “Common in brushy tracts of country. On Feb. 12-15 a half- dozen deer were shot in the Santa Cruz Mountains in Sonora. I shot a young buck April 2 in the Chiricahua Mountains which was shedding, having lost the greater part of its winter coat.”— Wi ..b. A very good account of this little deer can be found in Coues and Yarrow (I. c.), quoted from Rothrock. 2. Lepus alleni Mearns. ALiEN’s Jack Rapsir. Lepus allent MEARNS, Bull. Am. Mus. Nat. Hist. II, No. 4, 1890, p. 294. Rillito, Pima Co., Arizona. Represented by 13 fully adult specimens, taken at Fort Lowell, Jan. 5-21, by Price and Condit. ‘The -collector’s measurements from the fresh specimens may be summarized as follows: ‘Total length, 626 (600-700) ; tail vertebra, 70 (45-90); hind foot, 136 (130-145); ear, 160 (156-165). ‘Iwo specimens, both females, measure respectively in total length 680 and 700, but no others exceed 630, and none fall below 600, five ranging between 600 and 630. “This splendid hare is abundant about Tucson and in lower por- tions of the desert belt. It is found both on the gravelly hills bordering the Rillito at Fort Lowell, and on the immense mes- quite and Zarrea plains of Tucson. It is somewhat shy, and hard to secure, except with a rifle. One rarely comes upon it suddenly. I have never seen it start up with the quick rapid flight of Z. 202 Bulletin American Museum of Natural History. |Vol. VII, textanus. It has a slow, apparently awkward gait, but its leaps are long, and it gets over the ground with surprising rapidity. In color and habits it is so very different from any other American hare, the wonder is that it should have so long remained unde- scribed.” —W. W. P. 3. Lepus texianus eremicus 4//en. ARIZONA JACK RABBIT, Lepus texianus eremicus ALLEN, Bull. Am, Mus. Nat. Hist. VI, 1894, p. 347. Separates issued Dec. 7, 1894. Five Fort Lowell specimens and tiree Fairbank specimens (taken Jan. 11 to April 18), present the following measurements : Total length, 576 (565-625) ; tail vertebra, 78 (72-95) ; hind foot, 128 (123-138); ear, 134 (128-140). “The common Jack Rabbit is abundant over the entire region to about 7000 feet elevation. In the desert region about ‘Tucson, this species is somewhat supplanted by Zepus allent. In the White Mountain region they occasionally wander from the pinion belt up into the pines as far as Cooley’s Ranch.”—W. W. P. 4. Lepus sylvaticus pinetis A//en. Mountain Woop Hare. Lepus sylvaticus pinetis ALLEN, Bull. Am. Mus. Nat. Hist. VI, 1894, p. 348. Separates issued Dec. 7, 1894. Represented by 2 specimens from the White Mountains, as already described (1. c.). It differs from the Cottontail of the lowlands by its darker coloration, much smaller ears, and much more heavily clothed ears and feet. ““Two specimens only were taken by Mr. Condit in White River Cafion, in the White Mountains, at about 8000 feet eleva- tion. A few others were seen in the same locality among oak Scrubs. Wis, bas 5. Lepus sylvaticus arizonze Aden. ARIZONA SAGE Harr.—Represented by 22 specimens, taken as follows: Fair- bank, 1 @ ad., March 5 (Price and Condit); San Bernardino Ranch, 2 ¢6 ad., April 8 and 16, and 2 $9 ad., March 24-25 1895.| Allen on Mammats from Arizona and Mexico. 203 (B. C. Condit) ; Fort Lowell, 2 24 and 3 29, Jan. 8-18, and 2 99 and 344, March 21-25 (Price, Condit and Miller) ; Chiricahua Mountains, 6 99, April, June, July and August (Price and Miller); Huachuca Mountains, 1 2 juv., May 21 (Price and Miller). These specimens appear to be all referable to the large-eared Arizona ‘Cottontail... Winter specimens (January and March examples) are purer gray on the rump and more heavily lined with black than the worn summer specimens. The Fort Lowell series of ro adults (5 taken Jan. 8 and 5 taken March 21-25) give the following measurements: Total length, 357 (340-371) ; tail vertebrae, 42 (35-50) ; hind foot, 80 (76-88) ; ear, 73 (69-78). “Common over the entire region up to about 7000 feet (except in the White Mountains), and occasional in the Huachuca and Chiricahua ranges to 8500 feet elevation. About Fort Lowell it is exeeedingly common, a dozen or more often being seen during an evening’s walk. Common also about the deserted stables and ’ buildings of the post, furnishing fine sport for moonlight hunts,’ =o WR ae 6. Thomomys cervinus, sp. nov. FAWN-COLORED GOPHER. Above fawn-colored,' clearer on the sides, more or less obscured by dusky over the middle of the back ; below gray, the fur plumbeous at base with long whitish tips. Ears black, enclosed in a blackish area ; sides of the nose and muzzle blackish ; inside of cheek pouches entirely pure white ; feet dull whitish; tail thinly haired, pale grayish fawn color above, slightly lighter below. Measurements.—Type (Price Collection), total length, 228 ; tail vertebrie, 63; hind foot, 28. ‘Two other specimens measure as follows: Length, 228 and 263 ; tail vertebrze, 63 and 79; hind foot, 28 and 32. Type, No. *$2;5', Am. Mus. Nat. Hist., 4 ad., Phoenix, Arizona, Oct. 20, 1894; J. Diefenbach. The species is represented by three specimens, collected for Mr. Price by J. Diefenbach at Phoenix, Arizona, Oct. 30, 1894, and were not received for examination till May, 1895. They were not marked for sex by the. collector, but are apparently all males. 1 Ridgway, Nomen. Colors, pl. ili, fig. 22. 204 Bulletin American Museum of Natural History. |Vol. V1, Thomomys cervinus is very different in coloration from any phase of 7. fulvus | have ever met with. It is a large, pale form, about the size of Z. fossor and 7. aureus, but very different in color Fig. ra. Fig. 2a. Fig. 1. Thomomys cervinus. Type, No. 4a, 6 ad., Phoenix, Arizona. Natural size. Fig. 2. Thomomys fulvus. No. 44h, 6 ad.,San Bernardino Ranch, Cochise Co., Arizona. Natural size. from either. In fact, in coloration, it bears a very close resem- blance to Geomys lutescens from Phillips Co., Kansas, but the pelage is longer, coarser, and less glossy. In cranial characters it is allied to 7. aureus. The rostral portion of the skull is espe- cially broad and heavy, and the whole skull massive in comparison with 7’ fulvus (compare Figures 1 and 2), ~ aan 1895.| Allen on Mammals from Arizona and Mexico. 205 7. Thomomys fulvus (Woodh.). Arizona Goruer.—The series of Zhomomys numbers 112 specimens, all but three of which (as noted above) seem to be referable to 7. fudvus. The localities represented are as follows : Fairbank, Feb. 22 to March 14, 29 specimens ; San Bernardino Ranch, March 27 to April 22, 33 specimens ; Fort Lowell, March 8-17, 7 specimens ; Huasava Mountains, Sonora, June 25, 26, 3 specimens ; Huachuca Moun- tains, Feb. 20, 1 specimen ; Chiricahua Mountains, March 14 to April 23, and June 4 to July 20, 24 specimens ; Graham Moun- tain, July 19, 3 specimens ; White Mountains, July 28 to August 5, 8 specimens. They vary in age from half-grown young to very old adults; representing, as they do, a period of over six months, the variation in color and character of pelage is also very great. Many of them are in molt, and thus often represent two phases of pelage in the same individual. Aside, however, from differences plainly due to either age or season, there is a wide range of purely individual variation in both coloration and cranial characters. It thus becomes necessary to consider some- what in detail several of the larger series, as those from Fairbank, San Bernardino Ranch, Fort Lowell, and the Chiricahua Moun- tains, in connection with much material from other localities, numbering altogether nearly 200 specimens that seem referable to what is here called Zhomomys fulvus, including several speci- mens from San Francisco Mountain, the type locality of the species. EXTERNAL CHARACTERS.—(1) /atrbank Series.—Three rather young specimens, for the most part in their first pelage, are dusky brown, slightly tinged with fulvous gray, darker (in one specimen black) along the middle of the back, more or less fulvous on the cheeks and sides of the shoulders, and blackish below, the hairs slightly tipped with ashy fulvous, or extensively tipped with yellowish. Other more or less immature specimens are dusky yellowish brown, more strongly dull yellowish brown on _ the sides, and whitish below, from the long whitish tipping of the hairs. Others are similar above to the last, but are strong rusty or fulvous gray below. From this pelage the animal molts into that of the adult, as shown by several specimens in changing pelage. 206 Bulletin American Museum of Natural History. {Vol. V1, The adults are strongly yellowish brown, sometimes more or less rufescent, the intensity of the tint varying in different individ- uals, with generally a slight admixture of blackish tipped hairs along the middle of the back, increasing in some specimens so as to form a more or less broad median dorsal band. ‘Two specimens, both old males, of the Fairbank series, are everywhere intense glossy plumbeous black, except the feet (and in one specimen the apical fourth of the tail), which are whitish, and the inside of the cheek-pouches, which are pure white. These specimens are doubtless simply melanistic, and are the only melanistic examples I have met with in a series of hundreds of specimens of the genus Thomomys. (2) San Bernardino Ranch Series.—This, as regards the ages represented and the amount and character of the color variations, is almost an exact duplicate of the Fairbank series just described. The adults, however, average slightly more rufescent, several of them being strongly ferrugineous. (3) Fort Lowell Series —These are also in general very much like the Fairbank specimens. The adults, however, average a little paler, as though somewhat bleached or faded, a difference probably attributable to the fact that they were taken somewhat later in the season. (4) Chiricahua Mountains Series.—These differ quite strongly from the others in being darker, through a much stronger admix- ture of blackish tipped hairs, and the shorter and darker shade of the fulvous apical portion of the pelage. Several of the middle- aged specimens are quite dusky brown with a slight tipping of dark yellowish brown. ‘The young of this series, which includes a specimen not more than one-fourth grown, are not appreciably different from the young examples in the other series, as both the lightest and the darkest half-grown young are found in the present series. The specimens from the Graham and White Mountains corre- spond very closely in every respect with those from the Chirica- hua Mountains. For comparison with these, good series are available from Fort Verde and Bradshaw, Arizona, from Santa Ysabel and Dulzura, San Diego Co., Cal., and from San Pedro Martir, Lower Cali- - fe 1895.| Allen on Mammats from Arizona and Mexico. 207 fornia. Specimens from any of the series can be almost exactly matched by specimens from each of the other series, except per- haps the San Bernardino Ranch series, which differs as a whole from any of the others in being less grizzled and redder. In general effect there is scarcely any appreciable difference between the Bradshaw, Santa Ysabel, and San Pedro Martir series, and between these again and those from the mountains of southeastern Arizona, including also those from the Graham and White Moun- tains to the northward. Measurements.—Of this series of 109 specimens only 75 can be considered as sufficiently adult to be made the basis of compara- tive measurements. Females greatly preponderate in all of the series, so that out of the 75 specimens, of which measurements are given below, only 25 are males. The males average larger than the females, but old females often equal or exceed in size the smaller males. MEASUREMENTS (AVERAGES AND EXTREMES) OF 75 SPECIMENS OF Thomomys fulvus. Sex and) | Locality. No.of | Total length. | Tail vertebra. | Hind foot. Ear. Specim. | | Fairbank... .... 94 | 235 (217-264) Ti = (62-90) 32 (29-33.5) 7.2 (7 -8) Lhe oe 15¢ | 219 (195-245) 61 (49-71) 29 (28-31.5) | 7 (6.5-7) S. B. Ranch 53 226 (205-239) 67 (58-74) 27.2 (25-29) | 7 (6.5-8) Aare 152 | 203 (193-221) 58 (56-62) | 27 (26-28) 7 (6.5-8) Fort Lowell .... 24 | 223 (220-227) 63 (62-64). | 27.5 (27-28) | 7.5 (7 -8) aL 32 212 (210-214) | 62 (60-85) | 28 (25-30) 6 (6 -6.5) Chiricahua Mts 73 211 (201-218) | 62 (50-68) 28 (27-30) 6.7 (6 -7.5) 2 11g | 198 (171-225) | 58 (50-68) | 28 (25-31.5) 6.6 (5.5-7.5) White Mts..... 14 202 ( 58 | 28 7 “he ergata 62 | 194 (183-213) 59.3 (52-64) | 28 (26-30) 6 (6 -6 5) | | | | The Fairbank and San Bernardino Ranch series average larger than either of the other series. An examination of the skulls shows that these two series contain a larger proportion of very old specimens than the others, which latter happen to consist almost wholly of young and middle-aged adults. CRANIAL CHARACTERS.—A comparison of the skulls of these several series reveals no appreciable differences by which one set can be distinguished from the others. There is a wide variation 208 Bulletin American Museum of Natural History. (Vol. VU, in size, besides that evidently due to age, and striking differences in certain structural details. ‘The remarks which follow will be based wholly on the San Bernardino series, and mainly on the females of that series. Variation due to Age—TVhe youngest specimen is a_third- grown female. The skull is very short in proportion to its width, the shortness being due mainly to the comparative non-develop- ment, at this age, of the rostral and interorbital portions. The interorbital breadth is relatively very great, often actually greater than in the fully adult, and the zygomatic arches are relatively narrow. ‘The interparietal is also relatively very large. This is well shown by the following comparative measurements and figures (Figs. 3-5) : Total Pre-_ | Rostral | Mastoid Zygo- Inter- Interparietal. No. length. oa length. | breadth. I eeathn erie Width. oes 6803 2 juv. .| 29 17-5 8 16 19 aOR s 6 3 6799 @ juv...| 30 18.5 823i) alo 18 6 5 3 6794 2 juv..-| 33 21 nit 18 21.5 6 3-5 2 (OWfoy) 2 Ele cal) “Si7/ 24 12 18.5 24.5 6 3 a 6767 6 ad...) 38-5] 23-5) 12 27 25 6 2.5 3.5 6753 6 ad..-| 42 26.5) 14 21 26 6 Als Baa The skulls given above as ‘ adults’ are only middle-aged, there being no very old skulls in the series. The last one given is a somewhat older male from Fairbank. Figs. 3-5. Thomomys fulvus. Natural size. Fig. 3. No. 8145, @ juv., San Bernardino Ranch, Arizona. Fig. 4. No. 8429,9 ad., San Bernardino Ranch, Arizona, Fig. 5. No. $43, @ old ad., San Bernardino Ranch, Arizona. G7979 1895. | Allen on Mammals from Arizona and Mexico. 209 In old individuals the zygomatic arches are sharply angular both anteriorly and posteriorly, with the sides straight and parallel to the axis of the skull. In younger skulls they may be more or less convex, but are generally slightly more expanded posteriorly than anteriorly. In the young the interparietal is much larger than in the very old adults, its lateral borders becom- ing, with increased age, more or less overgrown by the encroach- ment of the parietals, as has already been noted in the case of Neotoma micropus (see this Bulletin, VI, 1894, pp. 233-246, Pl. iv). In young and middle-aged individuals the interparietal is usually quadrate and nearly twice as broad as long. In very old exam- ples it becomes more or less wedge-shaped, and longer than broad. in extreme old age sometimes little or no trace of it remains, it having become wholly buried. This, however, is much more frequently the case in 7. do¢te than in 7. fulvus. In several old examples from Fairbank it has the form of an obtuse wedge, which is apt to be more or less truncate in front, widening gradu- ally backward to very near the posterior border, where it widens rapidly so as to form on each side a narrow, pointed, latero- posterior angle. From middle age on, a slight temporal ridge is developed, which in old age becomes strongly marked on each side of the inter- parietal area. In none of the specimens, however, is there a single median sagittal crest. The superior border of the temporal muscle is outlined on the skull quite early in life by a slight, raised line, which later on becomes pushed nearer the median line by the continued deposition of osseus matter, till in old age the two lines are only from 1 to 2 mm. apart. They are generally parallel, and extend from the front of the brain-case to the occi- pital crest ; later they become continuous with the slightly raised edges of the interorbital area. As these ridges thicken and move toward the median line they encroach posteriorly upon the inter- parietal, the lateral edges of which become buried beneath them, thus greatly altering the shape of its visible portion. As already intimated, this process is frequently carried so far in 7. Jotte as to give rise to a well-defined sagittal crest, thus entirely conceal- ing the interparietal. [ June, 1895.| 14 210 Bulletin American Museum of Natural History. \Vol. V1, Sexual Variation—The skulls of males are generally larger than those of females of corresponding age, and more heavily ossified, but in other respects there seems to be no very apprecia- ble difference. Individual Variation.—In skulls of the same sex, and appar- ently of the same age, there is quite a range of variation in size, so that large females may exceed the dimensions of small males. Thus in old males from Fairbank the length of the skull varies from 38 to 42 mm., and the zygomatic breadth from 24 to 26 mm.; in old females from the same series the length varies from 36 to 39 mm., and the breadth from 22 to 24 mm. The width of the nasal bones, and correlatively the width of the rostrum, varies considerably in individuals of the same sex and age. But the most variable feature is the size and form of the interparietal, which may be twice or three times as large in some specimens as in others. While usually quadrate, and nearly as long as wide, it may be more or less convex on the posterior border, or, in rare cases, regularly convex anteriorly from the nearly straight posterior border. As these variations (see Figs. 3-8) occur in each of the large series at hand, and in about the same proportion, they cannot be considered as other than indi- vidual. In probably 75 per cent. the interparietal is distinctly four sided, with nearly straight outlines, except for a tendency Fig. 8. Figs. 6-8. Thomomys fulous. Natural size. Fig. 6. No. 8188, @ ad., San Bernardino Ranch, Arizona. Fig. 7. No, §#i%, ? ad. San Bernardino Ranch, Arizona Fig. 8. No. §442, 9 ad., San Bernardino Ranch, Arizona. 1895. | Allen on Mammals from Arizona and Mexico. 211 to slight irregularity on the front border. The variation from this is toward a convex outline in front, the convexity varying from a slight rounding of the antero-lateral corners to one involv- ing the whole of the lateral edges, resulting in a uniform convex outline extending to the posterior border, as in 7 /o/tecus. In case, under these circumstances, the interparietal is also small and narrow, it closely resembles the same bone in 7. doéle. Figs. g-11. Thomomys fulvus. Natural size. Fig. 9. No. 8139, 2 ad., San Bernardino Ranch, Arizona. Fig. to. No. 483, @ ad., San Bernardino Ranch, Arizona. 0 ° 6 F 5 : Fig. 11. No. #44 ad., San Bernardino Ranch, Arizona. The lower surface of the skull also presents much variation in details in specimens from the same locality strictly comparable as to age and sex. A single feature—the pterygoid hamuli—is here selected for illustration. As shown in Figs. 9-11, these vary in respect to the angle of divergence of the processes and in their conformation. In consequence of these variations it is almost impossible to point out any single cranial character that may be relied upon as absolutely diagnostic. ‘The rostral and interorbital portions of the skull are broader than in 7° do/¢@, and the general form of the skull is quite different in the two species. On the other hand, the rostral portion is less developed than in the 7. aureus group. “This gopher was the most generally distributed of any of the mammals taken during the expedition. It was found almost everywhere from Fort Lowell to the summits of the Huachuca, 212 Bulletin American Museum of Natural History. (Vol. VU, Chiricahua, Graham and White Mountains. It apparently does not hibernate at all. I have known it to throw up earth under several inches of snow. It was especially abundant on the sum- mit of Chiricahua Mountains during June and July. Ata par- ticular glade covered with grass and iris, often a half-dozen might be seen at once at nightfall raising their curious mounds of damp earth.’”—W. W. P. 8. Dipodomys deserti Stephens. DESERT KANGAROO RAat.— Two specimens from Sentinel, Maricopa Co. (J. Diefenbach, Dec. 20) are referable to this species. Another specimen from Phoenix (J. Diefenbach, Noy. 20) is intermediate in coloration between D. deserti and D. spectabilis, but in cranial characters is similar to the specimens from Sentinel. These localities carry the range of D. deseréz much to the east- ward of former records. The Phoenix specimen differs so much in coloration from the others, and also from true D. desertz, as to suggest that it may represent a strongly-marked local form of the desertt group. 9. Dipodomys spectabilis d/erriam. BANNER-TAILED KancGaroo Rar.—Represented by 20 specimens from Fairbank, Feb. 26-March 14 (Price and Condit); and 5 from San Bernar- dino Ranch, March 28-April 1 (B. C. Condit). All except 3 are fully adult. Of these latter one is nearly fully grown, and two are about one-third grown. ‘They are very similar in coloration to the adults, but the tail, though white at the tip, is not bushy. The adults are very uniform in coloration, allowing for the wearing off of the tips of the fur in a few of the specimens. ‘The white at the tip of the tail, however, varies in extent in different specimens from about 45 to 85 mm. The females average slightly smailer than the males, as shown by the following measurements of 14 males and 8 females. Males : Total length, 344 (330-363) ; tail vertebra, 198 (188-208); hind foot, 53 (49-56) ; ear, 17 (15-19). Females : Total length, 322 (302-345); tail vertebrae, 185 (168-205) ; hind foot, 51 (48-57); ear, 16.5 (15-17). BA: 1895. | Allen on Mammals from Arizona and Mexico. 213 “These beautiful Kangaroo Rats are pretty well distributed in colonies over the entire southern part of Arizona. They have hillocked towns not unlike those of Cynzomys, and well-beaten trails from one hillock to another. ‘The entrances, however, are horizontal, and usually enter the mound just above the level of the surrounding ground. This is no doubt a wise provision against rain, which often falls in terrific showers, and would otherwise flood the nest. One moonlight night at Willcox I had an opportunity of watching their habits. Secreting myself by a large hillock from which several trails radiated, I had not long to wait before I heard a slight noise on the gravel. It was a Rat approaching from another hillock, perhaps thirty yards away. It made low leaps of from one to several feet, and, as nearly as I could distinguish, ran, or alighted only on its hind feet. Several were sometimes leaping about the hillock at the same time. Some had ventured a dozen feet or more away, as if searching for seeds. During all the time I heard no sound of any kind, except a low chuckle uttered at intervals. They are difficult to secure with baited traps, but are readily caught in steel traps placed in the runways or entrances to their homes. ‘They breed early, for hali- grown young were caught March 1.”—W. W. P. 10. Dipodomys merriami J/earns. Merriam’s KANGAROO Rar.—This species is represented by 156 specimens, collected as follows: Fairbank, Feb. 22—March 12, Price and Condit, 93 speci- mens; San Bernardino Ranch, March 22—May 1, B. C. Condit, 33 specimens; Fort Lowell, Jan. 19 and March 8—April 20, L. Miller, 25 specimens; Phoenix, November 5—Dec. 12, J. Diefenbach, 5 specimens. The Fairbank and San Bernardino Ranch series are practically indistinguishable in respect to both size and coloration, but the Fort Lowell series averages appreciably smaller and more yellow. As usual with the Kangaroo Rats, the females average consider- ably smaller than the males. [For comparison the measurements of the several series are given separately, as follows: 214 Bulletin American Museum of Natural History. |Vol. VU, MEASUREMENTS (AVERAGES AND EXTREMES) OF 112 SPECIMENS oF Dipodomys merriamt. MALES. No. of| Locality. speci-| ‘Votal length. | Tail vertebra. | Hind foot. Far. ee mens. Fairbank..... .... 41 |246 (232-2641)| 139 (120-152) | 39.4 (86-42) 13.7 (13-15) S. B. Ranch...... 12 |248 (232-261) | 142.6 (129-154) | 38.4 (37-40) 14. (13-15) Ft. Lowell........! 12 | 283.5 (222-255) | 136.5 (132-155) | 36.6 (85-88) | 14 (12-15) _ FEMALES. Fairbank......... 30 | 238 (223-264) | 137 (124-150) | 38.2 (36-40.5)| 13.5 (12.5-15) S. B. Ranch...... 11 | 236 (222-248) | 140 (124-146) 38 (35-410) | 13.5 (13-14.5) Ft. Lowell........ 6 | 227 (215-246) | 131.6 (126-147) | 36.6 (86-38) | 13.7 (12.5-15) | 1 For one specimen the total length is given as 271. Young specimens, one-fourth to two-thirds grown, do not differ appreciably in coloration from adults. “This is the most abundant Kangaroo Rat in southern Arizona where it bears the same relation to the kangaroo rats as Perog- nathus obscurus does to the pocket mice. It apparently does not hibernate at all, as specimens were caught on the coldest and most stormy nights. Its burrows, placed anywhere in sandy soil, are often closed during the daytime.” —W. W. P. 11. Perodipus chapmani (d/earns). CHAPMAN’s KANGAROO Rar.—Of this species Messrs. Price and Condit collected 17 specimens at Fairbank, Cochise County, Feb. 22-28, the only locality at which they seem to have met with it. It was taken with Dipodomys merriam?, 32 specimens of which were collected at Fairbank between the same dates, and many others at the same locality later, as well as at other points in Cochise County. The two species greatly resemble each other in size and coloration, and are readily distinguished externally only by the presence of the rudimentary fifth toe on the hind foot of P. chapmant. Of these 17 specimens 14 are males and 3 females; all are practically adult except two, which are but little more than half- grown. The 11 fully adult males measure as follows: ‘Total length, 232 (223-247); tail vertebra, 127 (120-136); hind foot, 38 (36-40); ear, 14 (13-15). Three fully adult females measure: 1895. | Allen on Mammals from Arizona and Mexico. 215 Total length, 231 (227-237); tail vertebra, 123 (119-130); hind foot, 38.5 (38-39); ear, 14 (13.5-15). I also refer to this species two specimens from Fronteras, Sonora, taken May 15, by Mr. B. C. Condit. One is a nearly adult female, the other a nursling, apparently not more than a few days old. The young specimen is very dark—almost black, with a very faint fulvous tipping to the hairs on the flanks and across the shoulders. All the white markings shown in the adult are, however, present, and being pure white are very sharply defined against the blackish ground color. “This species was not uncommon at Fairbank, where it was found associated with Dipodomys merriami in the proportion of about one to three of the latter. They apparently live together, as specimens of each were caught from the same hole. “Mr. Condit shot at Fronteras, Sonora, May 15, a female carry- ing a young in its mouth during the daytime. This was the only specimen taken, besides those at Fairbank. So far as we could determine, its habits are identical with those of Dipodomys mer- aaa, WW. P. 12. Perognathus flavus Baird. YeLLow PocKET-MOUSE.— Of the 1o specimens representing this species, 4 were collected at Fairbank (Feb. 28—March 2, Price and Condit), and 4 at Fort Lowell (April 5 to May 9, L. H. Miller). Three very young speci- mens (about half grown) from the Chiricahua Mountains (July 4-21) are also referred here. “Quite common in fine sandy soil among bunches of sacaton grass at Fairbank. Its burrows, no larger than a little finger, usually ran horizontally into a small mound, and were often closed during the day with fine sand. It readily ate rolled oats, but was rather difficult to catch in our cyclone traps, owing to its small size and light weight. Mr. Miller found it not uncommon in the sandy fields about Fort Lowell. A single specimen was caught alive in a field at the western base of the Chiricahua Mountains, on July 4. It was said to be quite common, being often turned up in plowing.”—W. W. P. 216 Bulletin American Museum of Natural History. (Vol. VU, 13. Perognathus bimaculatus’ Merriam. ARIZONA PocKET-MOUSE.—This species is represented by 11 specimens, all from Holbrook, collected August 25-29, by Price and Diefenbach. Three are quite young, one being little more than half grown. In this specimen the general color above is pale yellowish drab, with a pale yellow lateral line, and a broad yellowish band on each side of the head, extending from the nose to the ear, the eye being at about the centre of this area. “ Found only at Holbrook, where the species was not uncom- mon on the sand flats along the Little Colorado River.”—W. Wie 14. Perognathus apache JMerriam. ApacHEe PockEr- mMousE.—Represented by an adult female taken at Fort Lowell, April 8, by Mr. L. H. Miller, and by two specimens from Hol- brook, taken August 26 and 27, by Price and Diefenbach. One is an adult female ; the other consists of the head and front half of the body and the skull, of unknown sex. “Two specimens were taken on sandy flats at Holbrook on August 26 and 27. They were found in company with Perognathus bimaculatus. A single specimen was caught by Mr. Miller in sandy soil at Fort Lowell, on April 8. These three are the only specimens obtained.”—W. W. P. 15. Perognathus obscurus Merriam. Brown Pocker- MOUSE. Perognathus obscurus MERRIAM, N. Am. Fauna, No. 1, p. 20, pl. iii, fig. 14 and pl. iv. Oct. 1889. Camp Apache, Grant Co., N. Mex. Perognathus pricei ALLEN, Bull. Am. Mus, Nat. Hist. VI, 1894, p. 318 (young). Oposura, Sonora. This species is represented by 168 specimens, collected at the following localities : Fairbank, Feb. 22 to March 15, Price and Condit, 57 speci- mens, of which 41 are males and 16 females. All are adult, as regards coloration, but about one-fourth are not quite fully 1 A large part of the specimens from Riverview, Utah, referred in a former paper (this Bulletin, V, p. 71) to P. afache prove on examination to be ?. démacudatus, both species being represented in the series. 1895. | Allen on Mammats from Arizona and Mexico. 217 grown, as shown by the measurements and the character of the skulls. Fort Lowell, March 8 to May to, L. H. Miller, 46 specimens, of which 29 are males and 17 females. Very nearly all are fully adult. San Bernardino Ranch, March 22 to May 4, B. C. Condit, 38 specimens, of which 32 are malesand 6females. All are adult as regards coloration, and with few exceptions also as regards size. Oposura, Sonora, May 30 to June 2, Bb. C. Condit, 16 specimens, of which 12 are males and 4 females; 11 are adult and 5 are young. Willcox, July 15, Price and Condit, 9 specimens, of which 3 are nursing females and 5 are young males; some of the latter are less than half grown, and all are in first pelage except one, which has begun to acquire the adult dress. Sentinel, Dec. 20, J. Diefenbach, 2, adult. From the foregoing it would seem that the young are born late in the season, apparently not till May or June, as the only very young examples, and the only females giving evidence of nursing young were the Oposura specimens taken the last of May, and the Willcox specimens taken July 15. It is also noteworthy that the number of males at all of the localities largely exceeds that of the females. The type locality of Perognathus obscurus is the southwestern corner of Grant County, New Mexico, but a few miles from San Bernardino Ranch, in the southeastern corner of Cochise County, Arizona. I am indebted to Dr. C. Hart Merriam for the loan of three ‘topotypes’ of PP. obscurus, taken in April and May, 1886, for comparison with the Arizona series. ‘They are slightly more fulvous than the average of the specimens from either of the localities mentioned above, but can be closely matched by exam- ples from either series, while the Fort Lowell specimens are practically indistinguishable as a series from the Grant County specimens. The young in first pelage are nearly uniform dark gray above, with a slight tinge of brownish, sparsely lined with blackish hairs ; below white, as in the adult. There is barely a trace of a very 218 Bulletin American Museum of Natural History. {Vol. VII, pale yellowish lateral line. Young adults are grayer and less fulvous than the fully mature individuals. The specimens from Oposura are all flat skins, so that the coloration is more condensed and the pelage apparently thicker than would be the case were the skins filled to life size. The young specimens from Oposura thus look very different from the young specimens from Willcox. On careful reéxamination of all the material it is evident that the young examples from Oposura, taken as the basis of my Perognathus pricei (|. c.) are not separ- able from P. obscurus. From the following summary of the measurements taken by the collectors from the fresh specimens, it will be seen that the Fort Lowell and Oposura (adult) specimens average slightly larger than those from the other localities; they are also more strongly fulvous. SUMMARY OF MEASUREMENTS (AVERAGES AND EXTREMES) OF 130 ADULT SPECIMENS OF VPerognathus obscurus. Sex and Locality. No. of Total length. Tail vertebrz. Hind foot. Ear. Specimens | Fairbank. . .. 32 4 180 (165-196) 87 (80-105) 23: (22 -24) 9 (8-10) se Bae 149 170 (160-178) 86 (82- 94) | 23 (22 -24) 8.6 (8- 9) Fort Lowell.. 27 ¢ 182 (170-200) 93 (88-110) 23 (21.5-24.5)) 8.7 (8 9.5) Bs x 149 173 (160-192) 86 (82-103) 22.6(21 —24) 8 3 (8-9 S. B. Ranch. 28 4 177 (166-190) 92 (81-107) 23 (2% -24.5)} 9 (8 9.5) re z 59 173 (164-183) | 91 (89- 96) 23 (22 -25) 8.7 (8-10) Oposura..... fies 188 (177-197) | 98 (78-105) 23 (22 -245)} 9 (8-10) Bo Ate 3¢ 177 (165-190) 94 (88-100) 22.5(21 -23) 9 (9- 9.5) Of the whole series of 130 specimens, 12 males and 4 females reach or exceed 190 mm. in total length ; and 16 males and 4 females reach or exceed too mm. in the length of tail vertebree. Fourteen males and 16 females fall below a total length of 170 mm. ; and 5 males and 6 females fall below 85 mm. in length of tail vertebrze. “This is the common Pocket-mouse of the region south of the Mogollon Mesa, where it outnumbers all the others, three to one. We found it especially abundant at Fort Lowell, Fairbank, Will- cox, San Bernardino Ranch, and at several points in Sonora. It was abundant at Fairbank as early as February 22, but as none re 1895. | Allen on Mammals from Arizona and Mexico. 219 were obtained at Fort Lowell in January, it is not unlikely that it ‘hibernates during the colder months. The holes sometimes descend perpendicularly into the ground, but usually enter hori- zontally into mounds heaped under mesquite bushes by wind ; during the hot weather it often closes the entrance with fine sand. These animals were caught readily with rolled oats. I often found seeds of various plants and mesquite béans in their peckers. —W. W. P- 16. Perognathus conditi 4//en. Conpir's PocKE?T-MousE. Perognathus conditi ALLEN, Bull. Am. Mus. Nat. Hist. VI, 1894, p. 318. (Separates published Nov. 7, 1894.) This species, as already noted (1. c.), was based on 3 specimens, taken at San Bernardino Ranch, March 23, 1894, by Mr. Condit. “On March 23 and 24 Mr. Condit caught two adult specimens of this brightly-colored Pocket-mouse in a boggy patch of ground thickly grown with sacaton grass. Later, on May 1, he obtained a third specimen in sandy soil among mesquite trees.” —W. W. P. 17. Microtus leucophzus (4//en). WHITE-BELLIED MEADOW MOUSE. Arvicola leucopheus ALLEN, Bull. Am. Mus. Nat. Hist. VI, 1894, p. 320. (Separates published Nov. 7, 1894.) The four specimens of a very white-bellied AZ/7crotus from Graham Mountain have already been recorded (1. c), and do not require further comment. “Four specimens were taken near the summit of Graham Mountain, at an elevation of about 10,000 feet, on July 18, 19. They were all found along boggy streams shaded with dwarf alders. There were no traces of runways in any of the surround- ing meadows.” —W. W. P. 18. Microtus alticolus (Merriam). Mounratn MEApow Mouse. Arvicola (Mynomes) alticolus MERRIAM, N. Am. Fauna, No. 3, 1890, p. 67, pl. v, figs. 1 and 2, and pl. vi, figs. 1-4. San Francisco Mountain, Ariz. Two specimens of a Microtus from the White Mountains (Aug. 3 and 8, B. C. Condit), an adult female and a young male, are pro- 220 Bulletin American Museum of Natural History. \Vol. VXI, visionally referred to this species. ‘The adult female measures : Total length, 168; tail vertebra, 50 ; hind foot, 21; ear, 17. “These specimens were obtained in the White Mountains at an elevation of about gooo feet. They were trapped among fallen logs along creeks where no runways were apparent. On a grassy meadow near the summit of the mountains, the well-beaten run- ways of an Arzico/a were abundant. Several hundred must have comprised the colony. ‘Though a dozen traps were set for them, not one specimen was taken.”—W. W. P. 19. Sigmodon minimus d/carus. Merarns’s Covron Rat, Sigmodon minima MEARNS, Proc. U. 5S. Nat. Mus. XVII, 1894, p. 130. Upper Corner Monument, New Mexico. Represented by two old males, taken at San Bernardino Ranch, April rr and May 9 (B. C. Condit), less than fifty miles west of the type locality of the species. These specimens measure : Total length, 246 and 241; tail vertebra, 95 and 99; hind foot, 28 and 25; ear, 19 and18. ‘They are slightly larger than the two speci- mens on which Dr. Mearns (1. c.) based the species, his examples, also “adult males,” measuring respectively as follows: ‘Total length, 223 and 223 ; tail vertebre, 94 and 91 ; hind foot, 28 and 27; ear, “above crown,” 14 and 12; “ above notch,” — and 16. ‘Two specimens of this species were taken by Mr. Condit at San Bernardino Ranch. ‘They were found in a boggy patch of ground, a half-acre or so in extent, thickly grown with coarse sacaton grass. Careful search failed to show traces of any others. Curiously enough, this species, which has just been described by Dr. E. A. Mearns from southern New Mexico, was found asso- ciated so closely with the new pocket mouse /. conditi that speci- mens of both were caught in the same spot.’”’—W. W. P. 20. Sigmodon hispidus arizonz J/earus. Arizona Cor- ron Rat.—The only locality represented is Fairbank, where a series of 12 specimens was taken by Price and Condit, Feb. 25 to March 15. Five are adult males and 7 are females, one of 1895. | Allen on Mammals from Arizona and Mexico. 221 which is quite young, and several others appear not to have quite reached mature growth, as shown by the subjoined measurements: No. of Sex. Yotal length. | Tail vertebra. Hind foot. Ear. specimens. 5 $ | 250 (238-300) | 111 (98-120) | 34.6 (33.5-36) 20 (19-21) 5 @ | 249 (232-277) 99 (go-104) | 33.6 (31.5-37) | 20 (Ig-2T) | “We found this species common in swampy localities along the San Pedro River at Fairbank. Nearly a dozen were trapped ina small patch of tules, where they had beaten runways in all direc- tions. Associated with them were Peromyscus arizone and Reith- rodontomys artzonensis. A Cotton Rat, probably of this species, is found at Igo’s Ranch, at the northend of the Huachuca Mountains. It was said to be common in a moist garden plot. However, I had no opportunity of visiting the place.”—W. W. P. 21. Neotoma mexicana Baird. Mexican Woop Rar.— Represented by 58 specimens, taken as follows : Fairbank, Feb. 25—March 3 (Price and Condit), 5 specimens ; San Bernardino Ranch, March 25-27 (B. C. Condit), 4 specimens ;_ Fort Lowell, March 7 (L. H. Miller), 2 specimens ; Oposura, Sonora, May 30 (B. C. Condit), 1 specimen ; Huachuca Mountains, Jan. 28, Feb. 12, and May 21 to July 3 (Price and Condit), 18 specimens ; Chi- ricahua Mountains, March 21, April 14, and June 17 to July 24 (Price and Condit), 18 specimens; Graham Mountain, Aug. 7 (Price and Condit), 1 specimen ; White Mountains, Aug. 1-4 (B. C. Condit), and Sept. 12, 7 specimens ; Showlow, Aug. 22 (B. C. Condit), 2 specimens. The series presents considerable variation in both external and cranial characters, but the variations are so inconstant, as shown by the large series from single localities, that it is difficult to con- sider them as not due to age and individual variation. Dr. Mer- riam has kindly sent me a number of specimens for examination, of what he considers to be true VV. mexicanus, as restricted in his later papers, and I see no impropriety in referring all of the 58 specimens in the Price Collection to this form. 222 Bulletin American Museum of Natural History. |Vol. V1, The coloration varies greatly with age and season. In fairly well- grown specimens the color of the upper parts ranges from a pale yellowish gray, more or less lined with black along the middle of the back, and with a decided wash of pale buffy on the sides (young adults) to a stronger yellowish gray quite heavily lined with black on the back, and with a stronger wash of buff on the sides (middle- aged specimens), and even to strongly rufescent brown above, heavily lined with black (old individuals). This variation is, how- ever, obviously (in part) individual and not wholly due to age. The whiteness of the lower parts varies with age, season and the length of the coat, the plumbeous basal zone being much broader when the pelage is fully developed than at earlier stages following the molt. In respect to the skull, the posterior branch of the intermax- illary usually extends considerably beyond the nasals, but in a small percentage of the skulls the intermaxillaries and nasals terminate on the same line, although in other respects the skulls are practically similar. The interparietal varies greatly with age, but to a rather less extent than in WV. micropus'—about as in WV. floridana. The teeth vary of course with age in respect to the character of the enamel folds, but also in specimens of corresponding age. Thus M! shows generally two deep sulci on the antero-internal border, of similar depth and character in comparatively unworn teeth. In old examples the anterior of these two sulci becomes more or less obliterated, sometimes wholly so, through the growth and wearing down of the tooth, while in very much worn teeth the other may also disappear. In M; the change due to growth and wear in the front border of the tooth is even more striking than in M1, the deep antero-internal sulcus seen in the young tooth becoming wholly obliterated in old age. In Mg the anterior loop is usually, or at least often, regularly convex on its anterior border, barely touching at its greatest con- vexity the tooth in front of it. In other specimens this anterior loop is flattened against the tooth in front, so that its front border is not only more or less flattened, but not unfrequently its antero- external border is developed into a slight angle, adding another (incipient) angle to the outer margin of the anterior loop. 1 See this Bulletin, VI, pp. 233-246, pl. iv. ey ‘male from Enterprise, Florida, (No. 1895. | Allen on Mammals from Arizona and Mexico. 223 Judging from the description, figure, and from topotypes of Neotoma albigula Hartley,’ from Fort Lowell, it is not separable from WV. mexicana. LN. pinetorum Merriam and JN. m. bullata Merriam are unrepresented in the present series. “Wood Rats were abundant over the entire country visited, from the summits of the Huachuca, Graham and Chiricahua Mountains to the lower desert regions. About Fort Lowell they were exceedingly abundant, having numerous nests among cactus beds, brush fences, and in willows along the Rillito. They appear equally at home among rocks, cactus, or oak brush, for, wherever we were, traces of Wood Rats were common.’’—W. W. P. NoTeE ON ECCENTRICITIES IN THE TEETH OF NEOTOMA.— Several specimens of Veofoma in the Museum collection present eccentricities that seem worthy of note. Neotoma californica Price—One of two topotypes of this species presents the following extraordinary deviations in Mg. It is an adult male (No. £{42) from Bear Valley, Cal. The last lower molar on each side has an extra enamel loop on the inner side, as though an attempt were made to reproduce the middle loop, normally developed in M; and Ms. A slight supernumerary cusp Is also seen on the outer side of M; and Mg of both rami, and two on the outer side of Ms. They are all merely incipient points arising from the cingulum, but are not without morphologic interest. (See Fig. 12.) Neotoma floridana (O7d).— A 7562) has normal dentition, except with respect to Mg of the right side, which has an extra circular loop of enamel on the outer side opposite the middle of the tooth. When worn down it might give the appearance of an additional loop Fig. 12. Neotoma californica. Lower molar series, three times on the outer side of the tooth, but has patural size. 1 Proc. Cal. Acad. Sci., (2) IV, pp. 156-160, pl. xii. 224 Bulletin American Museum of Natural History. [Vol. VII, Fig. 13. Neotoma floridana. Lower molar series, three times natural size. Fig. 14. Meotoma floridana. Left lower molar series, three times natural size. Fig. 15. Neotoma micropus. Left lower molar series, three times natural size. Wigs 17. - Figs. 16 and 17. Neotoma cinerea occidentalis. Left lower molar series, three times natural size. now the form of a flat-topped or trun- cated cone. In the corresponding molar of the opposite side there is a tendency to the same condition. (See Fig. 13.) Another specimen (No. 4888, 9 ad., Gainesville, Fla.) has a well-developed angle at the antero-internal border of the posterior loop of Mg (Fig. 14). In still another specimen (No. 4359, 4 ad., Han- cock Co., Miss.) a well-defined angle is developed at the antero-external border of the anterior loop of M3. A similar variation is seen in a speci- men of WV. micropus (No. 1234, ¢ ad., Rockport, Texas, Fig. 15). Less marked variations are not infrequent in JV. for7- dana, NV. micropus and WV. mexicana, as already noted in regard to the latter. In JV. cinerea occidentalis this aberration is frequently well marked, as shown in No. 2928 (Fig. 16), in comparison with No. zasz (Fig. 17). Fig. 17 may be con- sidered as representing the more usual or normal form. While these variations are in the main to be regarded as abnormal, they indicate tendencies to a more varied tooth-pattern, past or to come. 22. Onychomys torridus Cowes. Arizona Scorpion Mouse. — To this species are referred 43 specimens, collected mostly within about 50 miles of the type locality (Camp Grant, Arizona), as fol- lows : 26 specimens from Fairbank, Feb. 21 to March 14 (Price and Condit); 14 from Camp Lowell, Jan. 19 and March 8 to April 18 (Price and Miller); 3 from San Bernardino Ranch, March 27 and 31 and May 1 (B. C. Condit); and 1 from Phoenix, Dec. 10 (J. Diefen- 1895. | Allen on Mammals from Arizona and Mexico. 225 bach). Of the Fairbank series all but two are adult ; of the Camp Lowell series 8 are adult and 5 are about half-grown young; of the three San Bernardino Ranch specimens, the two March exam- ples are adult, and the May specimen is only about one-third grown. The adults are for the most part very uniform in coloration. The white tip to the upper surface of the tail usually occupies the apical fourth or third, but is occasionally almost wholly lacking. The lower parts are pure white (not “yellowish white, or an extremely pale buff or fawn,” as originally described from an alco- holic specimen), with more or less of the extreme basal portion of the fur pale plumbeous or ashy. The young are ashy gray above, more or less varied with blackish ; below as in the adults. The eight adults from Camp Lowell measure as follows : Total length, 148 (142-153); tail vertebre, 52 (47-55); hind foot, 22 (19.5-22.5); ear, 17 (15-19). The 24 adults from Fairbank average slightly smaller, as fol- lows: Total length, 144 (134-157); tail vertebra, 43 (40-49); hind foot, 22 (20-23.5); ear, 17.5 (16-r18.5). In total length 5 exceed 150 and 5 fall below 140; in length of tail only 3 exceed 45 and only 5 fall below 42; in length of hind foot 5 exceed 23 and 3 fall below 21; in length of ear only 2 exceed 18 and only 3 fall below 17. “This form appears to be abundant south of the Mogollon Mesa wherever there are sandy mesquite covered plains and river- bottoms. We found it abundant at Fort Lowell, Fairbank and Willcox. Itlives in holes under bushes and brush-heaps, and is partially carnivorous, for we frequently found the stomachs filled with scorpions, insects, and the hair and flesh of mice. ‘They would often drag off our traps containing small mammals. We sometimes found a trap containing a half-eaten mouse lodged in the opening of this animal’s burrow.’”’—W. W. P. 23. Onychomys leucogaster pallescens Merriam. DESERT Scorpion Mouse.—A series of 7 specimens, 4 of them not quite adult, taken at Holbrook, Apache Co., Arizona, Aug. 26-29, by Messrs. Price and Diefenbach, are referable to Dr. Merriam’s O. pallescens, which appears to be essentially a pale desert form [ June, 7895. | 15 226 Bulletin American Museum of Natural History. |Vol. VII, of O.leucogaster. The 5 oldest specimens (mostly ‘ young adults ’) measure as follows : Total length, 145 (135-151); tail, 40(37-45); hind foot, 21 (19.5-22.5); ear, 19 (16-20). “This species was found only at Holbrook. It was common on the sandy flats along the little Colorado, having holes in the sand heaped about bushes. It is a powerful little rodent, and was troublesome in carrying off our traps and their contents— White-footed Mice and Pocket-mice. They have a very peculiar musky odor.’”’—W. W. P. 24. Peromyscus eremicus (ard). Desert Mouse.— This species is represented by 7 specimens—2z adults from Fair- bank, Feb. 23-26; 1 adult from Fort Lowell, March 7; 2 adults from Phoenix, Dec. 12-14, and 1 adult and 1 nearly full-grown young from Oposura, Sonora, May 31. The young oné has a fluffy fulvous patch on each side of the abdomen, and is other- wise strongly suggestive of Hesperomys (Vesperimus) anthonyi Merriam (Proc. Biol. Soc. Wash., IV, 1887, p. 5), based on a series of immature specimens from Fort Apache, Grant Co., New Mexico. “Found sparingly at several places ; three or four specimens were trapped by brush fences at Fort Lowell, and in open fields at Fairbank. Mr. Condit found a few about the buildings at San Bernardino Ranch, and in fields below the town of Oposura.”— W. W. P. 25. Peromyscus auripectus 4//en. SitKy Ciirr Mouse. Sttomys auripectus ALLEN, Bull. Am. Mus. Nat. Hist. V, 1893, 75. Bluff City, Utah. Represented by 14 specimens from Holbrook, Apache Co., collected Aug. 26-29, by Messrs. Price and Condit. Two are quite young, 4 are adults, and the others ‘ young adults,’ still in more or less grayish pelage. Only two show any trace of the salmon-colored pectoral spot, usually present in adults. The 4 adults of the series measure as follows: Total length, 192 (184- 210); tail vertebrae, too (gt-107); hind foot, 23 (22-24); ear, 20,5 \(2o=2 0): 1895. | Allen on Mammals from Arizona and Mexico. 227 This species has the soft, silky pelage, and nearly the size and proportions of .S. evemicus, from which it is readily separable by its very hairy, heavily penicillate tail, and hairy heels, and when adult, by its lighter yellowish coloration above, and usually by the presence of a fulvous pectoral spot. “We found this form not uncommon among the sandstone ledges and cliffs along the Little Colorado River at the town of Holbrook. We caught them readily in traps baited with rolled oats or raisins. In some places they undoubtedly inhabited the nests of WVeotoma (sp.?).”—W. W. P. 26. Peromyscus rowleyi 4//en. RowLry’s WHITE-FOOTED Mouse. Sitomys rowleyi ALLEN, Bull. Am. Mus. Nat. Hist. V, 1893, p. 76. Nolan’s Ranch, Utah. To this species are referred 2 specimens from Showlow, and 2 from the White Mountains, Apache Co., Arizona. All are adult, and were taken respectively Aug. 22 and July 28. They meas- ure: Total length, 207 (198-210); tail vertebra, 106 (104-107); hind foot, 22.7 (22-24); ear, 20 (19.5-21). “Two specimens were taken at Showlow, just at the overlap- ping of the pine and juniper belts. 'wo were taken on White River, in the White Mountains, a few miles east from Fort Apache, fuiye27. ——W. W. P. 27. Peromyscus rowleyi pinalis (J////c,). | MiL.er’s WHITE-FOOTED MOUSE. Sitomys rowleyt pinalis MILLER, Bull. Am. Mus. Nat. Hist. V, 1893, p. 331. Granite Gap, Grant Co., New Mexico. This subspecies is represented by 132 specimens, of which 74 are from the Chiricahua Mountains, collected May 14 to July 2t, by Price, Condit and Miller; 40 from the Huachuca Mountains, collected Jan. 28 to Feb. 20, and May 21 to May 27, by Price and Condit; 11 from Huasava Mountains, collected May 24-27 by B. C. Condit, and 6 from Oposura, Sonora, collected May 30 228 Bulletin American Museum of Natural History. \Vol\. VU, by B. C. Condit. The adults are very uniform in coloration, but the immature specimens present every phase from the ashy gray young, washed strongly with black on the back, to the fully adult. The adults, however, vary. much in general size, in the relative length of the tail, and especially in the size of the ears, which, however, seem to keep pace with the general size in the increase with age from ‘ young adults’ to very old adults. A series of 46 adults from the Chiricahua collection measure as follows: Total length, 199 (185-225); tail vertebra, 98 (87—- 115); hind foot, 22 (20-24); ear, 19.6 (17-24). In total length 4 exceed 220, and 12 fall below 190 ; in length of tail 7 exceed rio, and 8 fall below 95; in length of hind foot 8 exceed 23 and 3 fall below 21 ; in length of ear g exceed 20 and 8 fall below 19. The smaller specimens are in many instances not fully adult in size, though practically so in coloration. The Huachuca series averages a little less, 16 adults measuring as follows: Total length, 192 (189-206); tail vertebra, 92 (S9- 109); hind foot, 22 (20-24); ear, 19 (17-21). In total length 6 only exceed 200, but only 1 falls below 190; in length of tail only 1 exceeds 1o5 and 5 fall below 95 ; in length of hind foot only 1 exceeds 23 and 4 fall below 21; in length of ear only 1 exceeds 20 and only 3 exceed 19. The Oposura series runs still smaller, ro adults measuring as follows : ‘Total length, 192 (184-195); tail vertebra, 94 (go-99); hind foot, 21 (20-22); ear, 19.5 (17.5-20). “This mouse is found in the region intermediate between | Peromyscus leucopus| rufinus of the higher altitudes and [ Peromys- cus leucopus | arizone of the plains. It was rarely found above 7500 or 8000 feet, and only at one place below 5000 feet. This was about 1o miles south of Oposura, Sonora, at an elevation of about tooo feet. There Mr. Condit found it not uncommon among brush fences and brush heaps along cultivated fields. This, with two specimens of P. eremicus, were the only forms of Peromyscus found in the region. It has all the habits of the genus, being found every- where, among rocks, brush heaps and iogs, and is also very troublesome about camp and in the houses of miners and pros- Peclons, VW yy Ar. , | Bex 1895.| Allen on Mammals from Arizona and Mexico. 229 28. Peromyscus megalotis (Merriam). LEAF-EARED Ciirr Mouse. Hesperomys megalotis MERRIAM, N. Am. Fauna, No. 3, 18go, p. 64. Black Tank, Desert of the Little Colorado, Arizona. Represented by a single adult male, taken at Holbrook, Aug. 29, by Messrs. Price and Diefenbach. The measurements from the fresh specimen, as recorded on the label, are: Total length, 185 mm.; tail vertebra, go; hind foot, 25; length of ear, 28; height of ear, 28. This specimen is slightly smaller than ?. megalotis, as described by Dr. Merriam from the Little Colorado Desert, but is otherwise similar. On the other hand, it differs from a series of 8 specimens of P. ¢ruet, taken near the type locality in New Mexico, in being less yellow and more tawny, in its much larger ears, larger size, and longer tail. The 8 specimens of P. ¢rwe/ measure as follows: Total length, 177 (165-184); tail vertebre, 87 (71-100); hind foot, 23 (22-23.6). The ears in ¢rvwei average fully one-fourth smaller than in mega/otis. “A single specimen of this huge-eared mouse was caught in sandstone cliffs along the Little Colorado at Holbrook on Aug. 29. Specimens of /?. auripectus were caught commonly within a few feet of this one.”—W. W. P. 29. Peromyscus leucopus SBOE SIS (Lacvie) SONORA WHITE-FOOTED MOUSE. Hesperomys sonortensis LECONTE, Proc. Acad. Nat. Sci. Phila. VI, 1853, p. RE on Ln provincia Sonorz ”’=Santa Cruz, Sonora.! Hesperomys sonortensts BAIRD, Mam. N. Am. 1857, p- 474 (in part ; only the Sonoran specimens); Mex. Bound. Surv. Zodlogy, 1859, Mam. p. 43. Hesperomys (Vesperimus) leucopus sonoriensis COUES, Proc. Acad. Nat. Sci. Phila. 1874, p. 179 (in small part—only the Sonoran reference); Mon. N. Am. Roden. 1877, p. 79 (Sonoran reference only). Sitomys americanus arizoné ALLEN, Bull. Am. Mus. Nat. Hist. VI, 1894, p> 321. ‘ In preparing the present paper it has seemed necessary to once more take up the question of Leconte’s Hesperomys sonoriensis, in consequence of the light thrown upon the general subject of the short-tailed mice of the middle region of the continent by the acs In Sonora, south-southwest [lege south-southeast] of Tucson. About lat. 31° 00’, long. ’ 122° o0'.”” Baird, Mam.N. Am.., p. 713. 230 Bulletin American Museum of Natural History. \|Vol. VU, very large series of these mice in the Price Collection, and from other sources, available for study in the present connection. As a result of this revision of the subject, | am led to consider that the form recently described by me as Szfomys americanus arizone is to be taken as the true sonoriensis of Leconte, using the name in a restricted sense for the short-tailed grayish brown form of Peromyscus of the open plains and semi-desert areas of southern Arizona and adjoining portions of northern Sonora. As is well known, the type locality of Hesperomys sonortensis I.econte was Santa Cruz, Sonora, and that the type itself was an immature example in the plumbeous phase of pelage, and thus not readily distinguishable from specimens of the short-tailed group of corresponding age from other localities further north, Hence, Professor Baird, in 1857, applied the name collectively to all of the short-tailed mice from the “Upper Missouri, and Rocky Mountains to E] Paso and Sonora.” In this he was fol- lowed by Dr. Coues in 1874 and 1877, and by authors generally till 1890, when Dr. Mearns’ “found that no less than five very distinct types are represented from the interior region of North America, viz.: a very dark arctic race ; a pale grayish form from the treeless plains of the north; a more reddish or cinnamon- colored race from the treeless regions of the south; a darker and browner southern alpine form; and a pallid race from the desert regions of California and Arizona.” Three of these had already re- ceived names ; to the other two new names were given, only one of which (Hesperomys leucopus deserticolus) requires consideration in the present connection. Dr. Mearns, however, redefined the other three, and the types of his diagnoses are before me. With Dr. Mearns’s material in hand, I am able to intelligently consider his work and allocate the forms he recognized. Unfortunately the name sonoriensis was restricted to the “darker and browner southern alpine form,” described soon after by Dr. Merriam as Hesperomys leucopus rufinus,’ and what was then and subsequently recognized as sonoriensis by other authors was re-named deserti- colus. I now propose to restrict sozortensis to the form I recently named arizona, and to let deserticolus stand for the “pallid race 1 Bull, Am. Mus. Nat. Hist., II, Feb., 1890, p. 284-287. 2N. Am. Fauna, No. 3, Aug., 1890, p. 65. 1895.| Allen on Mammals from Arizona and Mexico. 231 from the desert regions of California and [immediately contiguous desert regions of | Arizona,” which seems to be clearly separable from the sonoriensis (as now restricted) of southern Arizona and northern Sonora. My Sitomys americanus arizoné (now Peromyscus leucopus sono- riensis) was based on a series of 42 specimens taken at Fairbank, Cochise Co., Arizona, Feb. 22 to March 15, 1894, by Messrs. Price and Condit. ‘To the same form are referred 26 specimens, mostly immature, from San Bernardino Ranch, collected by Mr. B. C. Condit, March 21 to May 4. Also a specimen taken at Fort Lowell, Jan. 5; another taken at Willcox, Jwly 15; and another from Fronteras, Sonora, taken May 16, also by Mr. Condit. The Willcox specimen is very gray and faded; the Fronteras specimen is like many of the examples from Fairbank. Nearly all of these specimens came from within 30 to 50 miles of Santa Cruz, Sonora, the type locality of sonorzensis. “This mouse was abundant at Willcox, Fairbank and San Ber- nardino Ranch, having habits like those of \S. sonordensis |=P. leucopus deserticolus|, though at Fairbank some were trapped in boggy patches of tule.”—W. W. P. 30. Peromyscus leucopus deserticolus (J/earns). Ders- ERT WHITE-FOOTED MOUSE. Hesperomys leucopus deserticolus MEARNS, Bull. Am. Mus. Nat. Hist. II, No. 4, Feb. 1890, p. 285. Type, No. 1175, Am. Mus., 4 ad., Mojave Desert, California ; F. Stephens. Hesperomys leucopus sonortensis MERRIAM, N. Am. Fauna, No. 3, Sept. 1890, p- 66. (Only in part of previous authors. ) Vesperimus americanus sonoriensis ALLEN, Bull. Am. Mus. Nat. Hist. III, Aug. 1891, p. 302. To this form I refer a series of 18 specimens from Holbrook (Aug. 26-29), and 6 specimens from Showlow (Aug. 20-22). Two of the Holbrook specimens (Aug. 23) are in the light reddish phase of coloration characteristic of autumn and winter, of which others show slight traces. A few are in the dusky ashy pelage of the young, but the greater part present a brownish mouse-color tint, much like that of the winter pelage of sonoriensis. A series of 12 adults from Holbrook measure as follows: Total length, 154 mm.; tail vertebrz, 64 ; hind foot, 20; ear, 18. 232 Bulletin American Museum of Natural History. |Vol. VII, “This was the most abundant mammal on the sandy flats about Holbrook, where it was associated with Peroguathus and Onycho- mys. All three genera frequently have holes under the same bush. A few specimens were found at the edge of the pine belt below Showlow. This species was not found south of the great San Francisco or Mogollon divide.”—W. W. P. 31. Peromyscus leucopus rufinus (Merriam). ALPINE WHITE-FOOTED Mouse. Hesperomys leucopus rufinus MERRIAM, N. Am. Fauna, No. 3, 1890, p. 65, pl. iii, figs. 5-8. San Francisco Mountain, Arizona. Stitomys sonortensis ALLEN, Bull. Am. Mus. Nat. Hist. V, 1893, p. 74. (Not typical.) To this subspecies I refer all of the mountain races of the short- tailed Peromyscus represented in the present collection. Unfor- tunately the several series are not all comparable as regards season and condition of pelage. ‘They include (1) a series of 68 speci- mens from the White Mountains, taken by Mr. B. C. Condit, Aug. 2-18, and 4 taken Sept. 2-18; (2) a series of 25 specimens from the Graham Mountain, taken by Messrs. Price and Condit, July 18-19 ; (3) a series of 89 specimens from the Chiricahua Moun- tains, taken by Messrs. Price and Condit, June 11 to July 9; forming a total of 182 specimens. I would also now refer to the same form the large series (130 specimens) collected by Mr. Charles P. Rowley in the mountains of Colorado and New Mexico, which I recently referred (this Bulletin, V, 1893, p. 74) pro- visionally to Sz¢omys sonoriensts. ‘There are slight shades of differ- ence between the series from the different localities represented, but there is also such a wide range of individual variation in color, size and proportions, and such an endless and complicated varia- tion resulting from season and age, that apparently nothing is to be gained by attempting to recognize in nomenclature the slight average differences in coloration or other features that may possi- bly exist in the various more or less isolated mountain ranges of Arizona, New Mexico and adjoining regions. ‘This is at least my present view of the case, with some 600 specimens of the sonxoriensis group before me for examination. With larger series from these and numerous additional localities, collected throughout the year, 1895. | Allen on Mammals from Arizona and Mexico. 233 it might be possible to predicate slight shades of difference for each isolated area, but the practicability of attempting such fine discriminations must be left to future research, and more abund- ant and better material, for determination. The White Mountain series seems not to differ appreciably from specimens of vufimus from the San Francisco Mountains, the type locality of the subspecies. They are mostly immature or in changing pelage, but a considerable number have so far acquired the fall dress as to show satisfactorily the deep tawny brown char- acteristic of typical ,wfinus. A series of 18 fully adult specimens give the following measurements: Total length, 153 (144-164) mm.; tail vertebra, 61.7 (52-69); hind foot, 19.6 (18-20); ear, 18.3 (17-19). The adults of the Graham Mountain series are in worn, transi- tion pelage, and present, with few exceptions, a broad blackish dorsal area, with the rump and sides tawny brown, paler and more mixed with blackish than the White Mountain series, apparently a seasonal feature. A series of 16 adults average slightly larger than the adults of the White Mountain series, measuring as fol- lows: Total length, 159 (150-170); tail vertebra, 68 (65-73); hind foot, 21.7 (21-22.5); ear, 18 (17-20). This is, hence, a large form, and should the dark band along'the dorsal region prove a fairly constant feature at all seasons, would well merit recognition in nomenclature. But this does not seem probable, as one speci- men shows a narrow transverse line of tawny red hairs behind the shoulders, and another has the whole top of the head and nape red—remnants, evidently, of a tawny red pelage of earlier date. The Chiricahua Mountains series is quite similar to the Graham Mountain series; the adults are mostly in change, blackish along the median line of the back (but not so uniformly so, the black- ness of this area appearing often in patches), and of a paler tawny on the sides of the body and lower back than the White Moun- tain series. In size they are just intermediate between the White Mountain and Chiricahua series, 66 adults measuring as follows : Total length, 155 (142-170); tail vertebrae, 65 (53-75); hind foot, 21 (19-22.5); ear, tg (17.5—-20.5). A series of 16 specimens of rufinus from the type locality, as given by Dr, Merriam (N. Am. Fauna, No. 3, Aug., 1890, p. 66), 234 Bulletin American Museum of Natural History. |Vol. VU, measures as follows: Total length, 160 (150-170); tail vertebre, 68 (56-75); hind foot, 20 (19-21). A series of 20 adults from La Plata, New Mexico (altitude, 6100 feet), measures as follows: Total length, 153 (145-179); tail vertebrae, 68 (60-79); hind foot, 21 (19-22). For convenience of comparison, these measurements may be tabulated as follows : MEASUREMENTS (AVERAGES AND EXTREMES) OF 1360 SPECIMENS oF Peromyscus leucopus rupfinits. No.o : Locality. speci- | Total length. | Tail vertebre. Hind foot. Ear. mens San Fran. Mts.. 16 160 (150-170) 68 (56-75) 20 (19-21) White Mts...... 18 153 (144-164) 61.7 (52-69) —s- 19.6 eos 18.3 (17-19) Graham Mts.... 16 159 (150-170) 68 (65-73) 21.7 (21-22.5) 18 (17-20) Chiricahua Mts. 66 155 (142-170) 65 (53-75) 21 (19-22.5) 19 (17.5-20.5) La Plata, N.M.! 20 153 (145-179) 68 (60-19) 21 (19-22) 1 The apparently relatively longer tail in the La Plata series is probably due to difference in methods of measuring. From the above it appears that the White Mountain series averages a little smaller than the others, but it is geographically most nearly related to typical ,wfimus, as it is also in coloration, as nearly as can be judged from the material at hand. “This form belongs to high elevations, and was exceedingly abundant on the summits of the Chiricahua and Graham Moun- tains, where they were the only Svfomys obtained. In the White Mountains it was abundant from the summit down as low as 6500 feet, but in the Chiricahua Mountains it was not found below 8000 feet. It is found everywhere—in boggy flats filled with fallen logs, on bare, rocky hillsides, in thick brush—equally at home.’’—W, W. P. 32. Reithrodontomys megalotis (Aad). BiG-kARED Harvest Mouse.—Five specimens from Fairbank, March 2-14 (Price and Condit), seem distinctly referable to Baird’s 2. mega- Jotis, the type locality of which is not far to the southeastward of Fairbank. ‘Three of the specimens are adult, and give the follow- 1895. | Allen on Mammats from Arizona and Mexico. 235 ing measurements: Total length, 143 (141-146); tail vertebra, 66 (62-72) ; hind foot, 18.5 (18-19) ; ear, 14 (14-14). I also refer to this species a single adult male from San Ber- nardino Ranch (April 20, B. C. Condit), which differs from the others in being somewhat larger, but especially in having much larger ears. This specimen measures: ‘Total length, 150; tail, 74; hind foot, 19.5 ; ear, 17.5. “This species was found at Fairbank, in marshy places along ; y | g the San Pedro River, where five specimens were trapped, March 2145 —W. W.P. 33. Reithrodontomys fulvescens (4//en). Sonoran Har- vEsT MOUSE. Retthrodontomys mexicanus fulvescens ALLEN, Bull. Am. Mus. Nat. Hist. VI, 1894, p. 319. Reithrodontomys fulvescens ALLEN, ibid. VII, May, 1895, p. 138. There is at present nothing to add to the accounts already given (1. c.) of the three adult specimens from Oposura on which this species was based. “This species was taken by Mr. Condit, May 31, ten miles south of Oposura, Sonora, Mexico, in the valley of the Yaqui River. They were found along brush fences and shrubby mesquite trees.”’ 2 NY ge 34. Reithrodontomys arizonensis 4//en. CuiriCAHuA Harvest Mouse. Reithrodontomys longicauda ALLEN (nec BarrpD), Bull. Am. Mus. Nat. Hist. VI, 1894, p. 320 (in text). Reithrodontomys arizonensis ALLEN, ibid. VII, May, 1895, p. 134. The 5 specimens (of which 4 are adult) on which this species is based, are from the Chiricahua Mountains (July 7-9, B. C. Condit). At first they were provisionally referred to 2. longicauda of California, with which they have many points of relationship. The 4 adults measure as follows: Total length, 149 (145-152) ; tail vertebree, 78 (74-80) ; hind foot, 17 (16-18) ; ear, 14 (13.5-14). 236 Bulletin American Museum of Natural History. (Vol. VU, ‘‘ Five specimens of this species were trapped on Rock Creek, in the Chiricahua Mountains, July 7-8, at an elevation of about 8000 feet. ‘wo were in rocks and dry soil away from the bed of the creek, and the others were caught under logs and brush near the water.’’—W. W. P. AppDITIONAL Notrr ON REITHRODONTOMYS.—I received from Dr. C. Hart Merriam, just too late for notice in the preceding paper on the genus Rezthrodontomys (antea, pp. 107-143), some forty specimens of this genus, representing three species and various localities. Among them is a series of to specimens from Mason, Mason Co., Texas, and one or two specimens from Gainesville, Cooke Co., Texas. These localities are of special interest, as they indicate the probable continuous distribution of the genus southward throughout the greater part of Oklahoma, the Indian ‘Territory and Texas. . The specimens from Gainesville and Mason, Texas, seem dis- tinctly referable to the 2. dyche¢ group, and, judging from present waterial, are not even subspecifically separable from Kansas specimens. ‘The Mason specimens are rather small, but as most of them are more or less immature, their exact status may be left for future decision. ‘These localities thus extend the distribution of R. dychei from 200 to 500 miles south of its previous known range—from southeastern Kansas to west-central Texas, or to within about one hundred miles of the known northern limit of 2. mexicanus intermedius. 35. Mus musculus Z7zv. Housr Mousr.—Represented by 18 specimens: 1 from Holbrook, 1 from Showlow, 1 from Fort Lowell, 1 from Willcox, 6 from Fairbank, 1 from the Chiricahua Mountains, and 7 from Phoenix. “The House Mouse was found to be common in several locali- ties, as Fort Lowell, Holbrook, Fort Apache, Fairbank, Willcox, and any place where much teaming was done. A single specimen was caught at a house in the Huachuca Mountains. Three years before a wagon load of seed grain had been brought there, and of two house mice nesting in the grain one had escaped. The one I caught was in all probability the one that escaped,”—-W, W, P, | | 1895. | Allen on Mammals from Arizona and Mexico. 237 36. Cynomys arizonensis Wearns. ARrizoNA PRarRIE Doc. Cynomys arizonensis MEARNS, Bull. Am, Mus. II, No. 4, 1890, p. 305. Near Willcox, Cochise Co., Arizona. Represented by a single specimen ( 4 ad.) from the Huachuca Mountains, taken Jan. 28 (Price and Condit). “A single specimen was shot January 28, on the plain at the base of the Huachuca Mountains. It was a warm day after a cold rain, and the animals were scratching out their burrows, and feeding on the dwarfed grass roots. We saw about twenty, and, by the number of hillocks, estimated the colony to number about 200 individuals. To the next town east it was nearly a dozen miles. Old settlers know of a time when no Prairie Dogs could be found about the Huachuca Mountains. These people thought that the dogs had emigrated from northern Sonora, Mexico. In the Sulphur Spring and San Simon Valleys, Prairie Dogs are found in numerous colonies, especially about Willcox and on the plain along the east base of the Graham Mountain. They evidently de not hibernate at all during the winter. Cyzomys are found in large colonies on the Ash Fork plains north of the Gila Range. “The Cynomys found about Snowflake and Holbrook are prob- ably Cynomys gunnisont. In places large colonies were found, but unfortunately no specimens were obtained.”—W. W. P. 37. Anisonyx' (Otospermophilus) grammurus (.Suy). LINE-TAILED SPERMOPHILE.—Represented by 1 specimen from Fairbank (4 ad., March 1, Price and Condit); 3 from the Chiricahua Mountains (224,12, all adult, April 17 and May 2g, W. W. Price); 1 from Fort Lowell (Q@ad., March 18, L. H. Miller); 2 from the Huachuca Mountains ( 4 and 9 ad., June 18 and 21, L.. Miller); and r from the White Mountains (Cooley’s Ranch, Sept. 15). Total, 8 specimens. “This is the common ground squirrel of Arizona; it is found everywhere over the entire region up to about gooo feet. At Fort Lowell, during the summer of 1892, it was common in brush fences, and many had their dens under the roots of cottonwood and walnut trees. ‘They were injurious to the growing crops of 1 Cf. Merriam, Science, new Ser., 1, No. 1. p. 18, Jan. 4, 1805. 238 Bulletin American Museum of Natural History. [Vol. VU, the Mexican settlers along the Rillito. I did not notice any during my stay at Fort Lowell in January, 1804. It is probable that they hibernate during the colder part of the year, as the first specimens seen were on a warm day, Feb. 7, at the mouth of a cafion in the Huachuca Mountains. “At our camp in the Huachuca Mountains, during 1893, they were very troublesome. A few minutes after our leaving the cabin they would swarm down from the cafion sides and carry off everything that was not securely boxed—bread, pork, dried fruit and potatoes ; nothing came amiss to them. On our return they would scatter to the rocks, and for long after there would be a chorus of shrill chattering calls. At Showlow and Snowflake they were troublesome to the farmers, but were got rid of by poison. At Cooley’s they were quite common among rocks and about fences; Wea Wied 38. Anisonyx (Ictidomys) tereticaudus (ard). Rounp- TAILED SPERMOPHILE.—Represented by 13 specimens taken at Fort Lowell, by L. H. Miller, March 7 to April 30. Of this series 3 are males and ro are females; all are fully adult. The early March specimens show no signs of molting; the pelage above is rather short and close, but soft; below it is thinner, longer, much softer, dusky or blackish basally, and whitish at the ends of the hairs. The late April specimens have completed the spring molt. In these the pelage is everywhere short and close, slightly rufes- cent or of a pale cinnamon cast above, and clear silvery white below, becoming blackish as the hairs increase in length. The 3 males measure as follows : Total length, 238 (231-251); tail vertebra, 71 (65-78); hind foot, 33.5. (32-35); ear, 6. The 10 females measure: Total length, 243 (227-263); tail vertebra, 79 (70-92); hind foot, 35.2 (33-37); ear, 5.6 (5-6.5). “Found only at Fort Lowell, where they were abundant every- where. They are shy and in such color harmony with the soil that they might pass for a rare species upon casual observation. Throughout May and June, 1893, I had an opportunity for observing them at leisure. It was hard to come upon them unawares, but by secreting myself in bushes near their burrows, a. 2 1895.| Allen on Mammals from Arizona and Mexico. 239 I often saw them come out, ten or a dozen, one after another, and feed upon small seeds and mesquite beans. They would hurry silently away to their holes at the first noise. They are silent animals, rarely uttering an alarm ncte. The young are much less shy, and can sometimes be surprised away from their holes and caught in the hand. My companion had a pet one that ran at will about the rooms and fed greedily on raisins and rolled oats. It slept at night in the warm ashes of the fireplace.”—W. W. P. 39. Anisonyx (Xerospermophilus’) canescens (Merriam). Hoary SPERMOPHILE.—Represented by a single specimen from Willcox, the type locality of the species. It is an adult female, taken July 15, by Price and Condit. It gives evidence of having recently nursed young. It measures as follows : Total length, 220 ; tail vertebrzx, 64 ; hind foot, 34 ; ear, ro. “A female of this species was taken at Willcox, Arizona, July 15, in a thicket of mesquite bushes. Several others were noticed.” aw. WP. 40. Anisonyx (Xerospermophilus) spilosoma macro- spilotus (J/erriam).—Four specimens are provisionally referred to this subspecies, originally based (N. Am. Fauna, No. 4, 1890, p. 38) on specimens from Oracle, Pinal County, Arizona. Two, both adult males, are from Fairbank (Feb. 23 and March 11, Price and Condit) ; one, an adult male, is from the San Bernar- dino Ranch (May 4, B. C. Condit), and the other from the Chiricahua Mountains (May 4, W. W. Price). This last is indis- tinguishable from the Fairbank specimens ; the San Bernardino Ranch specimen has a slightly hoary tint, due perhaps to the in- coming post-breeding pelage. Hence in general effect it somewhat resembles the Willcox specimen, referred above to 5S. canescens. The four specimens measure as follows : Tail Hind foot. Ear. | Sex. | length | vertebrz. sf — e 2 ten ee 4 195 | 60 33 9 °° Oh HES beri 5 198 | 61 31 Chiricahua Mts..........- 5 210 64 pl -F | 8.5 5. BLL ee 4 217 | 72 35 8 a A i A al) Se ee 1 Cf. Merriam, Proc. Biol. Soc. Wash., Vil, 1892, p. 27, footnote. 240 Bulletin American Museum of Natural History. \Vol. VU, “This species is rather common about Fairbank among mesquite thickets in sandy soil. The animals are shy, and in habits are much like A. ¢ereticaudus. Their burrows are often placed at the roots of mesquite bushes, the beans of which form a large part of their food. Along the west base of the Chiricahua Moun- tains in the Sulphur Spring Valley are several large colonies. Here they have hillocked towns not unlike those of the Prairie Dogs. They can often be seen sitting upright above their burrows. A single specimen was taken from a small colony in the Sulphur Spring Valley. It is probable that the habitats of this form and that of A. cryptospilotus overlap, for only a level plain of 20 or 30 miles separates the two forms. Mr. Condit found a small colony at San Bernardino Ranch and collected a single specimen.” —W: W. P. 41. Anisonyx (Xerospermophilus) cryptospilotus (J7/er- riam).—A single specimen, ? ad., from Holbrook (Aug. 27, Price and Diefenbach) is provisionally referred to this species. It is very pale in coloration, with very faint whitish spots. ‘Total length, 216 ; tail vertebre, 65 ; hind foot, 31-5 ;_ ear; 0. “A single specimen was taken on the sandy alkaline plain bordering the Little Colorado River at Holbrook, August 28. No others were seen.” —W. W. P. 42. Anisonyx (Ammospermophilus') leucurus cinna- momeus (Merriam). WHITE-TAILED CHIPMUNK.—One speci- men, 4 ad., Holbrook, August 28, Price and Diefenbach. “We trapped a single specimen in the sandstone cliffs near Holbrook. It was in what I supposed to be a nest of Meotoma. No others were seen.” —W. W. P. 43. Anisonyx (Ammospermophilus) harrisii (4vd. & Bach.). HARRIs’s CHIPMUNK.—Six specimens, of which 5 are from Fort Lowell (Jan. 8-11, Price and Condit), and 1 from Phoenix (Dec. 12, J. Diefenbach). 1 Cf. Merriam, Proc. Biol. Soc. Wash., VII, 1892, p. 27, footnote. 1895. | Allen on Mammals from Arizona and Mexico. 241 “This species is rather common in the lower desert region of southern Arizona. It was taken at Fort Lowell on rocky hills, east of the Rillito, on January 8. Several specimens were seen, and a female containing nine embryos was obtained. On the cactus-covered plain stretching down to the Gila River from Graham Mountain, I found this species abundant on July 20. They were feeding on the seeds of the screw-pod mesquite, and one specimen shot had his cheek pouches distended with the shelled beans. Owing to the excessively hot weather no specimens mere preserved.”’—W. W: P: 44.. Tamias lateralis (Say). Sayv’s GrouND SQUIRREL.— Represented by 17 specimens, all taken in the White Mountains in August and September (Aug. 2-12, Price and Condit, 14 speci- mens; Sept. 4-13, Diefenbach, 3 specimens). All are adult except two, of which g are males and 8 are females. The August females are still in worn breeding dress, with traces of the incoming post-breeding pelage. The August males are somewhat advanced in molt, but in none is it more than half completed ; the September specimens have all completed the molt. Six fully adult males measure as follows: Total length, 270 (250-279); tail vertebra, 94 (80-109); hind foot, 41 (40-43) ; ear, 21.5 (20-23). In one specimen the tail vertebrae measure 109 mm.—11 mm. longer than in any other specimen in the series. Four old females give the following : Total length, 271 (255- 288); tail vertebra, 88.5 (85-92) ; hind foot, 41 (41-42); ear, 22 (21-24). “This species was common about Cooley’s Ranch, where they frequented rock piles, rubbish heaps and fallen logs ; some even had holes in the open woods. They were very tame, sometimes coming into my camp picking up crumbs. They do not resemble the Zamias proper in habits, but, being terrestrial, they are more like the small Spermophiles. Mr. Condit found them to be com- mon in the White Mountains to near the summit.’”—W. W. P. 45. Tamias dorsalis Baird. Gita CuipmunK.—Repre- sented by 105 specimens, 3 of which are from the Santa Catalina Mountains (Jan. 16), 2 from the Graham Mountain (July 19), [ June, 1895.) 16 242 Bulletin American Museum of Natural History. |Vol. VU, and 100 from the Chiricahua Mountains. Of the latter, 5 were taken March 29 to April 6, and the remainder May 31 to August 16. The sexes are not quite equally represented, there being 43 males and 57 females. Nearly all are adult. The measurements of 22 adult males and 28 adult females furnish the following summary : 22 males—total length, 224 (215-236) ; tail vertebra, too (go-110); hind foot, 34 (32-36); ear, 21 (19- 22): 28 females—total length, 233 (220-247); tail vertebrae, 103 94-114) ; hind foot, 34 (32-36) ; ear, 21 (19-23). Three males (14 per cent.) and ro females (36 per cent.) reach or exceed 235 mm. in total length ; 3 males and 13 females reach or exceed to5 mm. in length of tail vertebree. “T found Zaméas dorsalis in the Chiricahua, Santa Catalina and Graham Mountains, and in the lower parts of the region drained by the Salt and White Rivers, which drain into the Gila. On the r4th of January, I shot three specimens high up in the Santa Catalina Mountains. One was at an elevation of nearly 8000 feet, and close to a snow field. This was enough to show that the species does not hibernate except perhaps for a few weeks during a heavy snow-fall. In the Chiricahua Mountains I found it continuously after my arrival there on March 1g. It was com- mon from the scrub-oaks atthe base to the thick firs and aspens on the very summit, 10,000 feet elevation. In the Graham Range the species was common from the base to the summit. Two specimens were taken in fir woods at about 10,000 feet above sea level. In the open pine woods south of Fort Apache, I noticed this species several times. At Fort Apache they were abundant in the lava cliffs along White River, often venturing to the row of officers’ quarters, placed close to the bank of the river. On warm days in August I have sometimes seen three or four together, sunning themselves on the ridge of a deserted house. “Tn the Chiricahua Mountains, I had the ortoppunity to study them for several months. ‘They were generally distributed in rocks, brush fences, thick woods and brushy hillsides. They are rather shy animals, not commonly found in trees, as is Zamzas cinercicoll’s ; they have the usual chipmunk call. On warm days in June they were very abundant in Morse’s Cafion in the Chiricahua 1895.| Allen on Mammals from Arizona and Mexico. 243 Mountains. Often as many as ten or twelve could be seen at once, playing among the rocks near my camp. “One of the odd facts of distribution is that in the Huachucas the genus Zamias is entirely wanting, though the mountains in every particular appear to be as favorable a habitat as either of the other ranges mentioned.” —W. W. P. 46. Tamias cinereicollis A//en. San Francisco Moun- TAIN CHIPMUNK.—AIl of the 56 specimens were taken in the White Mountains August 6-20 and Sept. 2-19. Of this August’ series, 30 are males and 17 are females ; about one-half are fully adult, and the remainder immature, including a few less than half grown. The females average slightly larger than the males, as shown by the following summary of measurements: 16 males— total length, 217 (205-228) ; tail vertebrae, 96 (go-103) ; hind foot, 33 (32-36); ear, 19 (18-21): 12 females—total length, 224 (207-238); tail vertebrae, 99 (go-106); hind foot, 33 (31-34); €ar, 9.5 (17-21). Five (17 per cent.) of the males and 6 (30 per cent.) of the females exceed 223 mm. in total length ; 5 males (17 per cent.) and 4 females (25 per cent.) reach or exceed roo mm. in length of tail vertebrz. “Found only in the White Mountains and in the heavy pine timber about Cooley’s Ranch. Mr. Condit found it on the peaks of the White Mountains up to timber line. 7. czneretcollis is arboreal, and rarely seen on the ground or in rocks. It is an active species, and has a rather loud, sharp call. It is confined to the pine and fir zone of the San Francisco plateau, and reaches the White Mountains from the Mogollon plateau. Near Cooley’s the ranges of this species and that of Zwmazs dorsalis overlap, the former occupying a strip of country from 15 to 25 miles broad. - At Cooley’s nearly all the specimens taken were in oak trees, and they evidently feed largely on the acorns.”—W. W. P. 47. Sciurus hudsonicus mogollonensis M/carvs. MEARNS’S CHICKAREE.—This form of the Chickaree is represented by 8 1 The September specimens are not labeled as to sex. 244 Bulletin American Museum of Natural History. |Vol. VU, specimens, all adult females, from the White Mountains (Aug. 9- 12, B. C. Condit). They seem quite indistinguishable from the series of 12 specimens collected by Dr. Mearns in the San Fran- cisco Mountains, on which the subspecies was originally based. The White Mountain series measures: Total length, 322 (310- 336); tail vertebrae, 131 (126-138); hind foot, 51 (49-53); ear, 26.5 (24-28). “Abundant in the White Mountains above 7000 feet ; probably extends to the limit of the fir zone. A noisy species, feeding largely on the cones of Douglass fir.”—W. W. P. 48. Sciurus hudsonicus grahamensis 4//en. Mounr GRAHAM CHICKAREE. Sciurus hudsonicus grahamensis ALLEN, Bull. Am. Mus. Nat. Hist. eee 1894, p- 350. (Separates issued Dec. 7, 1894. ) There is at present nothing to add to the description (I. c.) of this form, based on 3 specimens from Graham Mountain, “This very restricted species is confined to the fir zone on the summit of Graham Mountain. ‘Three specimens were obtained in dense fir woods on Aug. 17 and 19. Others were heard chat- tering.”—W. W. P. 49. Sciurus aberti Woodh. Aperr’s SQuirreL.—Repre- sented by 6 specimens (12, 5@@, all adult), 5 of which were taken in the White Mountains, Aug. 1-8 and Sept. 17 (Price and Diefenbach), and 1 at Showlow. Four of the six measure as follows : — = 7 = = ] qa ee ———— = = = Sex. | Total length. Tail vertebra. Hind foot. Ear. é 500 | 238 65 41 % | 495 232 73 43 + | 515 | 235 | 75 43 2 | 521 | 229 62 45 2 | 498 . 221 63 41 “This handsome squirrel was common in the White Moun- tains, ranging from about 6000 feet up into the spruce belt to 1895.| Allen on Mammals from Arizona and Mexico. 245 about gooo feet. It was more abundant between 7000 and 8000 feet elevation. It has a loud ‘barking’ call and feeds on cones of Pinus ponderosa, and usually builds its nest of branches in some lightning-blasted tree.”—W. W. P. 50. Sciurus arizonensis Cowes. ARIZONA SQUIRREL.— Represented by a single worn specimen from Fort Apache, taken by Mr. Price. “ A single specimen was shot in pine and oak woods near Fort Apache on Aug. 20. It is probably found all through the lower pine zone, usually not overlapping the range of S. abert:.”—W. 1 5 51. Sciurus arizonensis huachuca 4//en. Hvuacuuca SQUIRREL. Sciurus arizonensts huachuca ALLEN, Bull. Am. Mus. Nat. Hist. VI, 1894, p. 349. (Separates published Dec. 7, 1894.) In addition to the 4 specimens on which this subspecies was based (I. c.), 3 have been since received, all being from the Hua- chuca Mountains. These additional specimens, taken June 9 and 1g and July 3, by Mr. L. H. Miller, are in worn summer pelage, but otherwise similar to those already described. “Common in the Huachuca Mountains from the highest peaks down to the base of the range, where I have found it feeding on walnuts in the cafions and ravines. During the sum- mer of 1893, with Mr. R. L. Wilbur, I found it abundant in Ramsey Cajon, which that year had a good crop of walnuts. Often we would see two or three in one tree feeding on the partially ripe nuts. A series of over 40 specimens was secured. ‘These squirrels often run upon the ground, but like best to jump from branch to branch. They have a call similar to that of Abert’s Squirrel. ‘They breed early, for by the middle of July we ob- tained young, nearly full grown.” —W. W. P. 52. Nyctinomus nevadensis (7. Ad/en). Nevapa Bat.— One specimen from the Chiricahua Mountains, as already recorded (this Bulletin, VI, 1894, p. 326). 246 Bulletin American Museum of Natural History. |Vol. VU, ¥ ‘““A single specimen, a female, was taken on‘the ridge of the Chiricahua Mountains at a small meadow called Fly’s Park, at about 9500 feet, on the evening of June 22. Another large bat, supposed to be of this species, was seen on the same evening. These bats were associated with large numbers of JV. braszliensts, and were flying from a dark cafon on the eastern slope of the range over the summit to the west.”—W. W. P. 53. Nyctinomus brasiliensis /. Geoff. Housrt Bar.—Rep- resented by 7 specimens, collected as follows: Huachuca Moun- tains, 9 ad., May 22 (Price and Miller); Chiricahua Mountains, 1 male and 5 females, all adult, June 19-23 (Price and Condit). The single male measures: Extent, 305 ; length, 102. The females range as follows: Extent, 298 (290-308); length, 97 (94-100). “These bats were abundant on the summit of the Chiricahua Mountains during June. From soon after sunset until too dark to see, a steady procession passed the summit from east to west. They had a rather steady flight, and did not apear to be feeding. Although they always appeared to fly from east to west, in the evening, it is likely they had a breeding place in the jagged cliffs on the east slope of the mountains, and returned there before daybreak, after feeding on the west slope. A single specimen was caught in a damp tunnel in the Huachuca Mountains on May 22. This species was exceedingly abundant at Fort Lowell through the month of May. Many specimens were taken in the cornice of the deserted hospital building.’—W. W. P. 54. Atalapha borealis (AZii//er'). Rep Bar. Represented by an adult female and two nursing young, taken in the Chirica- hua Mountains, June 27 (Price and Condit). “A nursing female with two young a few days old, was taken from the thick foliage of a peach tree at Wilgus P. O., at the west base of the Chiricahua Mountains, on June 26.”—W. W, P. 1 Cf. Rhoads, Am. Nat., XXVIII, June, 1894, p. 523 ; Reprint of Ord’s Zodl., 1894, App., p. 3- 1895. | Allen on Mammals from Arizona and Mexico. 247 55. Atalapha cinerea (Acaw.). Hoary Bar.—A specimen labeled : “Found on a wire fence, Huachuca Mountains, June 15, L. Miller,” consists of the complete skeleton and the hair, “A single specimen was found dead on a fence in Miller’s Canon in the Huachuca Mountains in May. This species was found to be not uncommon in the range of mountains during the summer of 1893.”—W. W. P. 56. Vesperugo hesperus (4. Aven). Picmy Bar.—One specimen, @ ad., Chiricahua Mountains, June 2 (W. W. Price). Expanse, 212 ; length, 77. “A single specimen was shot flying over an alfalfa field at the mouth of Rucker Cafion on June 2. A small bat, supposed to be this species, was one of the earliest to be seen evenings at my camp in Rucker Canon... They lived in cliffs on the caion side and flew high, with a wavering flight.”—W. W. P. 57. Adelonycteris fusca (Aeauw.). Brown Bar.—Repre- sented by 38 specimens, nearly all adult, and equaily divided as to sex. They were collected as follows : Chiricahua Mountains, May 31 to July rr (B.C. Condit), 12 males, 18 females = 30 speci- mens ; White Mountains, August 7-16 (W. W. Price), 6 males and 2 females. With the exception that the younger specimens are darker and smaller than the others, there is very little variation in coloration or size, there being no appreciable sexual variation in color, and very little in size. The females average slightly larger than the males, as shown by the following summary of measurements: 17 adult males, expanse, 326 (300-345); total length, £14 (204-120): 17 adult females, expanse, 334 (310-354); total length, 118 (111-125). “Abundant everywhere, from the desert region about Fort Lowell, to the summit of the Chiricahua Mountains, 10,000 feet above sea level. A specimen taken May 30 contained several foetuses. At Fly’s Park, on the summit of the Chiricahua range, bats of this species were the first to appear after sunset. They had homes in the dense forest of firs which walled one side of 248 Bulletin American Museum of Natural History. [Vol. VII, the glade, and with Z. xoctvigans appeared to be the only bats that lived on the summit of the Chiricahua Mountains. At the saw mill on Rock Creek, on the west slope of the Chiricahuas, every evening these bats, singly and in companies of fours and fives, were seen flying down the cafion. At Cooley’s Ranch this bat was abundant, outnumbering all the others. They appear to roost in all conceivable places, in cliffs, barns, hollow trees, tun- nels and culverts.” —W. W. P. 58. Lasionycteris noctivigans (Zecontc). SILVERY-HAIRED Bar.—Three specimens, Chiricahua Mountains, June 11 and 23 (Price and Condit). “Three specimens were taken and several others seen at Fly’s Park, on the summit of the Chiricahua Mountains. They inhab- ited the forest of firs, and at nightfall came into the glade to feed.”—W. W. P. 59. Vespertilio nitidus 4. Allen. Catirornia Bar.— Four specimens— @ ad., Chiricahua Mountains, June 29 (Price and Condit). Forearm, 38; 3d metacarpal, 35; total length, 100; expanse, 260. White Mountains, August 8 (W. W. Price), 2 males, measuring respectively : forearm, 37 and 37; 3d metacarpel, 32 and 33; total length, 84 and 85; expanse, 245 and 250. “Three specimens of this species were taken; one at the saw millon Rock Creek, in the Chiricahua Mountains, on June 29, and two at Cooley’s Ranch, in the White Mountains.” —W. W. P. 60. Vespertilio melanorhinus Merriam. BLACK-NOSED Bat.—Two specimens are provisionally referred to this species, namely, an adult male taken in the White Mountains, August 2 (B. C. Condit), and a male (apparently young) taken at San Ber- nardino Ranch, May 4 (B. C. Condit). The White Mountain specimen is of the same golden-brown color above as the type ; the other is darker, more resembling the ordinary dark phase of V. nitedus. White Mts..... 4ad. Forearm, 32.5 Total length, 84 Expanse, 2.40 5, B. Ranech?, 2) Suv. * Br.5 = —_ is 1895. | Allen on Mammals From Arizona and Mexico. 249 “Mr. Condit obtained a single specimen of this species in one of the buildings at San Bernardino Ranch on April 15, and on July 29 a second specimen under a stone in the White Mountains at an elevation of gooo feet.”—W. W. P. 61. Vespertilio evotus AH. Allen. Lonc-EARED Bat.— One specimen, Huachuca Mountains, 4 ad., July 3 (L. H. Miller). Forearm, 35; 3d phal., met. 1, 33; thumb, 6; total length, 85 ; expanse, 237 ; height of ear from crown (in dry skin), 14; height of tragus, 8. I also refer to this species an adult male from the White Mountains (August 8, W. W. Price), which resembles the other in size, color, and in all external features except that the tragus in each ear is defective, being square, hollowed at the top, and only about 2mm. long. Thisstrange condition may be due to malformation or to mutilation in life, as the two stumps are not quite symmetrical in outline, the upper border of the tragus hav- ing a different outline in the two ears. “Mr. Miller obtained a single male from the thick branches of an oak in the Huachuca Mountains, and I collected one specimen at Cooley’s Ranch on August 15, which flew into the house after dark attracted by the light.” —W. W. P. 62. Vespertilio lucifugus Zecente. BLUNT-NosED Bat.— Two specimens, from Cooley’s Ranch, White Mountains, are here referred to what has usually passed current as Vespertilio lucifugus, of which species it seems to be a western form, the type locality of Leconte’s V. /ucifugus being South Carolina. ‘““A single specimen was shot at Cooley’s Ranch, flying over a small pond by a house, feeding.” —W. W. P. 63. Antrozous pallidus (Zeconte). Pate Bar.—One speci- men, ¢ ad., Cooley’s Ranch, White Mountains, Aug. 15 (W. W. Price). “A single specimen was taken at Cooley’s Ranch in the White Mountains on Aug. 15. Bats supposed to be of this species were rather common flying high over the pines about the ranch _ build- 250 Bulletin American Museum of Natural History. [Vol. VU, ings. ‘They appeared early in the evening, but flew high and were difficult to secure.”—W. W. P. 64. Procyon lotor hernandezii (/Vag/.). BuAack-roorEeD Raccoon.—Two specimens, an adult male and an adult female, taken at La Noira (at head of Santa Cruz River, ten miles north of the Mexican town of Santa Cruz), Feb. 1 (Condit and Morgan), are provisionally identified as above. ‘They represent the pale southern form of P. Zofor. ‘The measurements of these two speci- mens are as follows : Total length, 2, 808, 2, 815; tail vertebre, 3:5 280, 2.3285 hind.footy.d ,/120, 94 me0;) ent oy 62a eeae “Raccoons were common in willow thickets along the Rillito Creek at Fort Lowell, and about Fairbank on the San Pedro River. The tracks of a few were seen along the streams at the base of the Huachuca Mountains, and a male and female were taken from a hollow oak at the International Line just south of Huachuca Mountains. On the night previous there had been a light fall of snow, and the animals were easily tracked to the oak. I did not see any signs of them in the Chiricahua Mountains, and old settlers informed me that they were not found in the range. At Cooley’s they were destructive to growing corn, pulling down the stalks, and eating the soft ears. ‘The Apache Indians are in many places compelled to guard their fields during the corn season on account of the ravages of this pest.”—W. W. P. 65. Mephitis estor Merriam. ARrizoNA SKUNK.— Two specimens —a very old maie and an old female—from Fairbank, taken respectively Feb. 27 and March 5 (Price and Condit) are referred to this species. ‘Chey present extremes of variation in color, the male having the principal part of the dorsal area, in- cluding the upper surface of the tail, white, with the underfur from the shoulders posteriorly dingy gray. There is also a narrow white lateral line, and a median band of white on the ventral surface, broad over the pectoral region, narrower and somewhat interrupted posteriorly. The lower surface and apical portion of the tail is somewhat mixed with black, white prevailing. ‘The usual frontal white stripe is reduced, however, to a narrow line. ee 1895.| Allen on Mammats from Arizona and Mexico. 251 The female is entirely black, except for a frontal stripe of white, and a white lateral line, very narrow anteriorly but widen- ing posteriorly where it forms a broad band. ‘The tail is black, with a small white terminal pencil, and much white at the base of the hairs, increasing in extent proximally, where many wholly white hairs are intermixed. These specimens measure respectively : Total length, 2, 545, @ , 682; tail vertebre, $ , 268, 2, 376; hind foot, 6, 68, 2 ,60; ear, 4, 34, 2,29. The skulls measure: Total length (front of base of incisors to posterior border of occipital condyles), 4, 66, 9 , 61; greatest zygomatic breadth, 4, 44,2, 39. In both the teeth are well worn, but more so in the male. In this specimen the tail is abnormally short. Since the above was written three additional specimens have been received from Fort Lowell, two of them taken Jan. 9g and 16 (Price and Condit), and the other March 18 (Price and Miller). These measure as follows : Orig. No. Sex. Total length. | Tail vertebra. Hind foot. Ear. 382 2 685 330 64 30 374 5 675 355 65 28 2004 3 630 280 72 2 The first two are without skulls; the skull of the other (No. 2004, dad.) measures 69 by 44. In this specimen the back is white with a narrow band of black posteriorly, and the tail is white at the base, along the sides, and at the tip. No. 382 ( 9 ad.) has a lateral white stripe running from the ear to the base of the tail, very narrow for the anterior third of its length, with a narrow broken white line above it, at the shoulders. ‘There is a well- developed frontal stripe, but no white on the nape or anywhere on the dorsal region between the white lateral bands. ‘The tip of the tail has a long white pencil, and there is a tuft of white hairs on either side of the lower surface of the tail at its base. The other specimen (No. 374, ¢ad.) has the usual frontal stripe, a broad white nape patch, continued posteriorly as far as the shoulders, and ending in a point. The rest of the body and tail 252 Bulletin American Museum of Natural History. (Vol. VU, are entirely black, except a few white hairs (about ten) at the tip of the tail, and a small amount of concealed white at the base of the tail hairs for the entire length of the tail.’ “Two specimens of this species were taken at Fairbank during February and March. ‘lhe species was common at Fairbank, and often during the night carried off many small traps containing kangaroo rats and mice. Specimens were also taken at Fort Lowell and in the Catalina Mountains. It is probably distributed over the entire region.” ——W. W. P. 66. Spilogale gracilis Merriam. LirrLe Srripep SKUNK. —Represented by two adult females from the Huachuca Moun- tains, taken Jan. 28 (Price and Condit). They measure respec- tively as follows: Total length, 325 and 338; tail vertebrae, 125 and 116; hind foot, 38 and 4o; ear, 28 and 26. HPA 6 I'wo specimens were trapped in a meat house at a ranch near my camp in the Huachuca Mountains in January. I obtained evidence of the occurrence of the Little Striped Skunk at many other places, but saw no other specimens.”’—W. W. P. 67. Bassariscus astutus® (Zicft.). RING-TAILED Cat.—One specimen, ¢ ad., Huachuca Mountains, Feb. 1 (Price and Con- dit). Measurements: Total length, 720 ; tail vertebrae, 345; hind foot, 68 ; ear, 50. “ A single male was caught in a trap at my camp in the Hua- chuca Mountains, Jan. 31, 1894. This species is rare in the Huachucas, though a few are killed every year by the miners and wood-choppers. ‘They sometimes come into the houses, and when young can be tamed, and are as playful as kittens. In the Chiri- cahua Mountains a single specimen had been killed several years previous to my visit, the only case of its capture of which I could find evidence.” —W. W. P. 1 For further notes on the variability of the Skunks of Arizona referred to Wefhitis estor, see Mearns (this Bulletin, III, pp. 258-262,) and Allen (this Bulletin. VI, pp. 194-106). 2 Mr Rhoads has recently proposed (Proc. Acad. Nat. Sci. Phila., 1893, pp. 413-418—sepa- rates dated Jan. 27, 1894) to separate *‘ the Bassarisks of Northern Mexico and the United States’’ from the true 2. astutus of southern Mexico, under the name of Bassartscus astutus flavus. mA 1895. | Allen on Mammals from Arizona and Mexico. 253 68. Urocyon cinereo-argenteus scottii (A/carns). Scorr’s Fox. Urocyon virginianus scottii MEARNS, Bull. Am. Mus. Nat. Hist. III, No. 2, 1891, p. 236. Pinal County, Arizona. Two specimens, as follows: An adult female, Fairbank, March 1 (Price and Condit). Measurements: Total length, 925 ; tail vertebra, 420; hind foot, 127; ear, 76. An adult male, Cooley’s Ranch, White Mountains, Sept. 4 (J. Diefenbach). Measure- ments: Total length, 906; tail vertebra, 363; hind foot, 121 ; ear, 79. “Scott’s Fox was seen over the entire region, but only two specimens were taken, one in March at Fairbank, and one at Cooley’s Ranch in September by Mr. Diefenbach. They were heard howling nearly every night at my camp in the Huachuca Mountains during the summer of 1893.”—W. W. P. 69. Lynx baileyi Merriam. PLATEAU Lynx.—Represented by two specimens : 4 ad., Huachuca Mountains, Feb. 1 (Price and Condit); @ad., Fairbank, March 12 (Price and Condit). The Huachuca specimen measures: Total length, 770; tail vertebrze, 155; hind foot, 165; ear, 80. The Fairbank specimen measures: Total length, 847; tail vertebra, 147; hind foot, 172 ; ear, 86. “Wild Cats were not uncommon over the entire country. Their tracks were seen on the summit of the Chiricahua Mountains, 10,000 feet above sea level. In the Huachuca Mountains a large male was caught ina trap in the day-time. Another was shot from a willow tree at Fairbank.”—W. W. P. 70. Felis concolor Zinn. PanrHEer; Mountain Lion.— One skull, 2 ad., Huachuca Mountains, Feb. 16. “The ‘Mountain Lion’ is restricted to the brushy and timbered mountains of the entire region. Occasionally this beast travels across the valleys from one range to another. One was seen on the San Pedro River above the town of Fairbank in February. It killed a colt in a pasture, and was tracked by dogs a dozen 254 Bulletin American Museum of Natural History. |Vol. VU, miles eastward into the Mule Mountains. In the Huachuca Mountains this animal is common. On Feb. 16, at nightfall near the summit of the range, two lions came mewing about the door of a miner’s cabin. The man shot through the door, killing one, a gaunt female. The next day he threw the skinned carcass a short distance from the house. During the night the other lion came and ate nearly the whole of it ; on the following evening the animal again returned, uttering a low peculiar cry. The miner wounded this one, but it escaped into the thick brush. In com- pany with the man I trailed the beast some distance through the snow, but we finally lost the track. The man kindly gave me the skull of the female he had killed. In the Chiricahua Mountains lions are exceedingly troublesome to the raisers of colts and don- keys. In some cafions horse ranges have become nearly depopu- lated by the ravages of this animal. Just before my arrival in Rucker Cafion a lion killed a mare weighing over 1500 pounds. Mr. Condit found the tracks of this animal at timberline on the White Mountains.”—W. W. P. I1I.—List oF MAMMALS OBSERVED IN THE REGION, BUT OF WHICH NO SPECIMENS WERE SAVED. By W. W. PRICcE.! 1. Covore. Canis latrans Szy.—Abundant over the entire region. Scarcely a night passed that bands were not heard howl- ing, or their tracks seen in the neighborhood of our camps. 2. Gray Wotr. [Canis lupus nubilis (Say). ? Canis lupus mextcanus (Linn.).|—This animal is the terror of the cattle and sheep men. A full-grown wolf is strong enough to pull down a cow, and stories are rife among the cattlemen of a band attacking and killing the strongest steer. It is found over the entire region, though more especially in the mountainous parts. We saw it on several occasions during our stay in the country. {! In some instances Mr. Price, in the following list, omitted to supply scientific names, or used names recently supplanted by others; these | have supplied or changed, as the case may have required, changes from the manuscript being indicated by inclosing the names in brackets. = ACTAG| 1895. | Allen on Mammals from Arizona and Mexico. 255 3. LONG-EARED Fox. [Vulpes macrotis MMZerriam.|—This fox is not uncommon on the San Simon Plain east of the Chiricahua Mountains, judging from the reports given me by the cattlemen. I heard of one specimen being taken at Fort Lowell previous to my arrival. I saw what I supposed to be a fox of this species early one morning while riding from the Chiricahua Mountains to San Bernardino Ranch. | There are two Arizona specimens in the Museum Collection —one from Tucson, collected by W. E. D. Scott, and one from near Maricopa, collected by Dr. E. A. Mearns.—J. A. A.| 4. Buack Bear. [Ursus americanus /a//as|.—Bears are found in all the mountainous and wooded regions of both Arizona andSonora. At Rucker Cajfion, in the Chiricahua Mountains, they were quite common during March and April. They had evidently left hibernation and were migrating. Bands of three or four, judging from the tracks, frequently passed through the canon. They were common in the White Mountains during August, where several were seen in the glades digging for roots and bulbs. Mr. Condit killed one on Aug. 4. 5. SILVER-TIPPED Bear. Ursus horribilis >—This huge bear is said by the natives to inhabit all the mountains, but this needs verification. So far as I was able to learn, only one ‘silver tip’ had been killed in southern Arizona in recent years. The skin of this one is now in the possession of Mr. J. H. Slaughter, owner of San Bernardino Ranch, and was killed by one of his men near Guadaloupe Cation. 6. Sorex, sp. ?.—A Shrew undoubtedly inhabits the fir belt of the principal mountain ranges. Dr. A. K. Fisher obtained two specimens on the summit of the Chiricahua Mountains near running water. I have seen its tracks on the Graham and the White Mountains. On one occasion I caught a tail of one in my trap. 7. WeasEL. Putorius feet elevation in the Huachuca Mountains during 1893, and from ?—A weasel was taken at gooo 2 56 Bulletin American Museum of Natural History. |Vol. V1, casual observation I supposed it to be P. braszliensis frenatus. The odor of weasels was noticed in both the Chiricahua and White Mountains, but no specimens were seen. 8. Bapcer. [Taxidea taxus berlandieri (Zaird).|— Badgers are common on the plains of the whole region. One was shot in 1893 at Fort Lowell, and is now in the collection at Stan- ford University. They are even found as high as Cooley’s Ranch, in the White Mountains. g. SporteD Cat. Felis, sp. ?.—A spotted cat has been seen about the Chiricahua Mountains on several occasions, and I saw a Mexican who had a saddle-bag made of a skin of one taken — near Guadaloupe Cajion. 10. Beaver. Castor canadensis Aw//.—The Beaver is still to be found along the San Pedro and Gila Rivers. On the headwaters of the San Pedro, in Sonora, a colony of a dozen or more had their lodges up to 1893, when a trapper nearly exter- minated them. All the streams in the White Mountains have beaver dams in them, although most of the animals have been trapped. tr, Pattip Muskrat. Fiber zibethicus pallidus d/earns. —Muskrats are found in the San Pedro River at Fairbank, and presumably at other points. A muskrat was common in Showlow Creek at Showlow, where were many trails leading from a pond up into an alfalfa field bordering it. Although we set traps for them, we did not secure any. 12. Sciurus, sp. ?.—A large Red Squirrel is rare in the Chiri- cahua Mountains, where I heard of it on several occasions through the settlers. According to them the animal is found in very diversified situations. A pair lived in 1893 in dense fir woods at the head of Rucker Canon in the southern part of the range. In Morse’s Cation, in the central part of the range, in 1892—'93, they were not uncommon at a low elevation, feeding on the cones of Pinus edulis. 1 searched diligently in both of these localities, ™% 1895.| Allen on Mammats from Arizona and Mexico. 257 but no traces of them could be found. However, in dense fir woods on the summit of the range, I found gnawed cones on several occasions, but I did not see the animals. Dr. A. K. Fisher, of the Agricultural Department at Washington, who was camping near me on the summit, had the good fortune to secure a single specimen in deep fir woods on June 17.. Mr. Condit saw a large red squirrel in the pines on the north slope of the Mogol- lon Mesa near Showlow on August 22. He is familiar with the Red Squirrel of the Eastern States, and thought it was that. It could not be captured. 13. AnTELorpE. Antilocapra americana O7d¢—Antelopes are still to be found on the plains of most of the region. Several bands were found along the bases of the Huachuca and Chiri- cahua Mountains. The most we saw in any band was twelve— a very different story from that told by old settlers of bands of hundreds, which in the early days trampled down the grass like sheep. We also found them in the juniper belt of the north slope of the Mogollon Mesa. 14. BLACK-TAILED Deer. [Dorcelaphus hemionus “a/.’]. —Still to be found in the foothills and ravines of the lower mountain ranges. They prefer a rather open country with oak woods. ‘They were formerly exceedingly abundant, but, like the Antelope, will soon become practically extinct. Some few bands still live along the west slope of the Huachuca Mountains. At a ranch house we saw some very fine antlers which had been taken during the fall of 1893. 15. Erx. Cervus canadensis 4rx/.—So far as we could learn this animal is now confined to a small area in the higher 1 [Cervus hemionus Rar., Am. Month. Mag., I, Oct., 1817, p. 436. Mule Deer of the Upper Missouri region. ‘ Cervus auritus WARDEN, Descrip. statis. hist. et pol. des Etats-Unis de l’Amer., Sept., V. 1820, p. 640. The Mule Deer of Lewis and Clark. In the English ed., 1819, I, p. 245, and IIT, p. 172, it is mentioned simply as the Mule Deer. Cervus auritus DESM., Mamm., II, 1822, p. 443. From Warden, as above. Cervus macrotis Say, Long’s Exped., II, 1823, p. 88. Rafinesque, on the basis of Le Raye’s brief description of the Mule Deer of the Upper Missouri region, gives, under the name Cervus hemionus, a fair diagnosis of the Mule Deer of the early explorers of this region, to which Sayin 1823 gave the name Cervus macrotis. There is also no question of the pertinency here of the name Cervus auritus given by Warden in 1820, over which, however, Rafinesque’s name has three years’ priority.—J. A. A.| [August, 1895.] LF 258 Bulletin American Museum of Natural History. [Vol. VII] White Mountains. Several were seen, and a fine male was shot at about gooo feet elevation, on August to. They feed in the dense fir woods and glades which clothe the upper slopes of the moun- tains. Mountain SHEEP. [Ovis cervina Desm.'|—Not uncommon on the bare rocky spurs of the Santa Catalina Mountains, where they were seen during 1894. Several were killed in the fall of 1893 by an Indian hunter, and the meat sold to settlers at the foot of the mountains. Some are also said to be found on the rocky, eastern flanks of the Chiricahua Mountains, but I found no positive evidence of their occurrence there. They are said to be found in the White Mountains, but none were seen there by our party. [! The proper specific designation of the Big-horn or Mountain Sheep has long been in doubt. In 1817 this animal was called by Cuvier Ovzs montana, he at the same time wrongly attributing the name to Geoffroy. Although there was a prior Owzs montana, given by Ord in 1815 to the Rocky Mountain Goat, this name passed current for the Big-horn till 1880, when Alston (Biol. Centr.-Am. Mam., 1880, p. 111) revived for it Ovzs cervina Desm., giving, however, not the place and date of its first publication. Mr. Rhoads has since reverted to the subject (Reprint of Ord’s Zoél., 1894, p. 25), and, among other things, says: “Shaw (Nat. Misc., XV, t. 610) figured and described this species under the name Ovwzs canadensis, but this work, being without any date whatever, the name is unavailable. though it probably has priority over any other.’’ Very recently Mr. C. Davies Sherborn has published in the ‘Annals and Magazine of Natural History’ (April, 1895, PP. 375 376) a short paper ‘On the Dates of Shaw and Nodder’s “ Naturalist’s Miscellany,’’’ from which it appears (granting that the work was published at the uniform rate supposed, of which proof is lack- ing), that plate 610 showld have been published in December, 1803. If there was any delay, even of but a few weeks, the part containing this plate could not have appeared till early in 1804. On the other hand, there is no question of the publication of Desmarest’s name Ovis cervina in 1804. There can be a difference at most of but a few months in the publica- tion of two names. Obviously the name having a positive date should have preference. (See A. O. U. Code of Nomenclature, Canon XLX, second paragraph under * Remarks.’ Biddulph (P. Z. S., 1885, pp. 682-684), in 1885, supposing O. cervzZna Desm. to date from 1818, adopted Shaw’s name canadensis, the date of which he gives as ‘* in or about r1804.”’ The history of this interesting case may be briefly presented as follows: Belier de Montagne, ¥. GEorrroy, Ann. du Mus. d’Hist. Nat., II, 1803, pp. 369-363, pl. 60. Canada, lat. 50°, long. 115’ =Rocky Mountains, in southern part of present Alberta, Can. (No technical name given.) Ovis cervina DesMareEstT, Nouv. Dict. d’Hist. Nat., XXIV, 1804, p. 5. Based exclusively on the above, the species here first receiving a scientific name. Ovis cervina DESMAREST. Nouv. Dict. d’Hist. Nat. (nouy. éd.), XXI, 1818, p. 553. Same in substance as the last, and is the reference usually cited. Desmarest here errone- ously cites ** Ovzs montana Geoffr.,”’ and gives the plate as “ pl. 40” instead of pl. 60, and omits the page reference. “ Outs cervina DESMARETS’’ RAFINESQUE, Am. Month. Mag., I, Oct., 1817, p. 436. Correctly attributes the name to Desmarest, but cites the date (by typographical error) as “ 1614” instead of 1804. Ovis cervina ALSTON, Biol. Centr.-Am. Mam., 1880, p. 111 (ex Desmarest, at 1818). “Outs canadensis SHAW, Nat. Misc., XV, pl. 60’’ (no date: about Jan., 1804). Apparently based also on the Berlier de Montagne of Geoffroy. (1am unable to verify this refer- ence; judging from contemporary literature, Shaw’s name, description and figure were doubtless based on Geoffroy, as above cited.) Ovis canadensis BIDDULPH, P. Z. S., 1885, p. 683 (in text). From Shaw, as above. Ovis montana Cuvier, Régne An., I, 1817, p. 267. Cites Geoffroy, and wrongly attributes to him the name Ovwzs montana —J. A. A.] Article VII.—LIST OF MAMMALS COLLECTED IN THE BLACK HILLS REGION OF SOUTH DAKOTA AND IN WESTERN KANSAS BY MR. WALTER W. GRANGER, WITH FIELD NOTES BY THE COL- LECTOR. By J. A. ALLEN. During the season of 1894 the Paleontological Expedition from the Museum, under the direction of Dr. J. L. Wortman, was accompanied by Mr. Walter W. Granger, an assistant in the Museum, who was sent into the field to utilize the opportunities for field work afforded by the Expedition. Mr. Granger gave most of his time to collecting the small mammals of the regions visited, but also collected many birds and reptiles. The mam- mals obtained number about 600 specimens, representing 42 species. The first stop was at Pendennis, Lane County, Kansas, where three days were spent, namely, May 8-10. Here 23 small mam- mals were obtained, representing 7 species, one of which (Veo/oma campestris Allen) proved to be new. The objective point of the Expedition was the Bad Lands of the White River Miocene, on the Pine Ridge Indian Reservation, South Dakota. Here Mr. Granger collected for a month (May 16 to June 14) in the early part of summer, and again for ten days in August (Aug. 18-27). In July he left the Expedition for a trip into the Black Hills, stopping en route at a ranch on Spring Creek from June 19 to July 4; also at a ranch on the Cheyenne River from July 5 to July 13; and on Squaw Creek from July 20 to July 23. Custer, in the Black Hills, was reached July 24, where Mr. Granger remained till August 9, when he moved camp to Hill City, some twenty-five miles north of Custer, where he spent three days, returning thence to join the Expedition again in the Bad Lands. Here work was continued till August 28. About a week in the early part of September was again spent in the Black Hills, at the abandoned tin mine known as Glendale. [259] 260 Bulletin American Museum of Natural History. [Vol. VII, Later about six weeks (Sept. 14-Nov. 2) were spent at Long Island, Phillips County, Kansas, where zodlogical collecting was carried on incidentally in connection with field work in paleon- tology. Although the material here under notice was gathered in part at quite distant localities, it has seemed best to combine the results of Mr. Granger’s work into a single consecutive list, giy- ing also nominal lists of the species obtained at each of the principal localities. The following descriptive account of the localities visited is based on notes kindly furnished by Mr. Granger. Pendennis, Lane Co., Kans.—I\n the prairie region of west- central Kansas. Most of the collecting was done in the ‘ cafions,’ from ten to a hundred feet wide and thirty to forty feet in depth. A few wild currant bushes and other small shrubs grow along the bottom of these cafions. Here the following species of mammals were obtained : ; Perodipus richardsoni. Reithrodontomys dychei nebrascensis. Perognathus paradoxus. Peromyscus leucopus texanus. Neotoma campestris. Spermophilus tridecemlineatus pallidus. Onychomys leucogaster. Long Island, Phillips Co., Kans.—Northern border of central Kansas. ‘The following species were obtained here : LLepus campestris. Sciurus niger ludovicianus. Lepus melanotis. Spermophilus tridecemlineatus pallidus. Lepus sylvaticus bachmani. Cynomys ludovicianus. Geomys lutescens. Scalops aquaticus argentatus, Mus decumanus. Mephitis mesomelas. Mus musculus. Spilogale interrupta. Peromyscus leucopus texanus. Corral Draw, Pine River Indian Reservation, South Dakota.— In the Bad Lands, between the Cheyenne and White Rivers, at the southeastern base of the Black Hills. Altitude about 3500 to 4ooo feet. The country is rough and broken, consisting of alter- nating buttes and cafions, cut in gray clay, interbedded with occasional layers of sandstone. Some of the cafions are quite deep, their sides formed of ledges of sandstone, and with water- worn caves in the intervening beds of clay. "ee 1895. | Allen, Mammats from the Black Hills Region. 261 Corral Draw is one of the many ‘draws’ or valleys that lead from the interior of the bad lands down to the Cheyenne River, a distance of about ten miles. These draws are beds of creeks, which are dry except after heavy rains. The vegetation is gener- ally scanty. A few of the higher buttes are flat-topped, and their level summits are well covered with a good growth of grass, cacti, sunflowers and other coarse plants. Sheep Mountain, the most prominent of these buttes, is covered with low cedars, which also grow in clumps on the slopes of some of the other buttes. In Corral Draw the soil is sandy and supports good grass. Cotton- woods extend about half way up the draw from the Cheyenne River. Spring Creek, S. Dak.—Spring Creek rises in the Black Hills and runs into the South Fork of the Cheyenne River. After leaving the Hills it passes through fertile prairie lands its entire length. It is bordered by boxelder, cottonwoods, plumb thickets, willows, wild currants, and rank weeds and grass. The specimens labeled as from Spring Creek were taken at a ranch seven miles from its entrance into Cheyenne River. Cheyenne River, S. Dak—The specimens labeled as from Cheyenne River were taken at the mouth of Spring Creek, and hence well out from the Hills. The species obtained here were the same as those from Corral Draw, with the following in addition: Perodipus richardsoni, Corynorhinus townsendt, and Taxidea taxus. Vhese doubtless also occur at the former locality. Sguaw Creek, Custer Co., S. Dak.—A small creek just in the edge of the pine forests of the Black Hills. Altitude about 3000 feet. The cafon through which the creek passes is wooded with aspens, willows, boxelders and other deciduous trees. ‘The few mammals obtained here belong distinctively to the Black Hills fauna. Custer, Black Hills, S. Dak.—Altitude 5500 feet. In the pine forests of the Black Hills. The collecting here was done in one of the numerous small parks near the town of Custer. This little park was about a mile and a half long by half a mile wide, and 262 Bulletin American Museum of Natural History. |Vol. VII, is drained by French Creek, which passes lengthwise through it. It is surrounded by pine forests. The land in the park is now mostly under cultivation. ffill City and Glendale Mine.—These two localities are practi- cally the same, as regards elevation and surroundings, as Custer. The South Dakota mammals are found to fall rather sharply into two groups, those from the Bad Lands (Corral Draw, Spring Creek, and Cheyenne River) and those from the Black Hills (Custer, Hill City, Glendale, and Spring Creek). The two cate- gories compare as follows : Black Hills. Lepus sylvaticus grangeri. Thomomys talpoides. Zapus princeps. Neotoma grangeri. Peromyscus leucopus arcticus. Microtus longicaudus. Microtus insperatus. Evotomys gapperi brevicaudus. Fiber zibethicus pallidus. Arctomys dacota. Spermophilus tridecemlineatus pallidus. Tamias quadrivittatus borealis. Sciurus hudsonicus dakotensis. Sorex forsteri. Putorius longicaudus. Bad Lands. Lepus sylvaticus nuttalli, Lepus campestris. Thomomys talpoides. Perodipus richardsoni. Perognathus paradoxus. Perognathus fasciatus. Erethizon epizanthus. Neotoma rupicola. Peromyscus |. nebrascensis. Microtus haydeni. Cynomys ludovicianus. Spermophilus t. pallidus. Tamias minimus. : Adelonycteris fusca. Vespertilio ciliolabrum. Corynorhinus townsendi. Taxidea taxus. The above lists are of course not. exhaustive for the localities treated, and more of the species are common to the two regions than these comparative lists indicate. It 1s interesting to note that there are several representative or parallel forms, according in coloration and in other features with their respective surround- ings, as strikingly illustrated in the genera Lepus, Peromyscus, Neotoma, and Tamias. 1895.| Allen, Mammals from the Black Hills Region. 263 A few species are included of which no specimens were obtained, these being given on the authority of Mr. Granger. They are mainly the larger game and fur-bearing animals of the Black Hills region, respecting which Mr. Granger gives interesting informa- tion.’ The collection contained a number of forms believed to be new, and as such were mostly described in the preceding volume of this Bulletin (Vol. VI, pp. 320, 322-326, 346). They are as follows : Lepus sylvaticus grangeri.” Neotoma rupicola. Neotoma campestris. Microtus insperatus. Neotoma grangeri. Sciurus hudsonicus dakotensis. 1. Cervus canadensis Zrx/. E.x.— The Elk has been extinct in the Black Hills for several years, but the numerous antlers which are to be seen at nearly every ranch show that it was recently not uncommon.’—W. W. G. 2. Dorcelaphus hemionus (/a/.). Mute Derr ; BLack- TAILED DrER.—‘ Numerous in the Black Hills. About extinct in the Bad Lands.’”’—W. W. G. 3. Dorcelaphus virginianus macrourus (/a/.)... WuiTeE- TAILED DEER.—“ Two White-tailed Deer came to an oat field near camp one morning at Custer. They were the only ones I saw.’—W. W. G. 4. Antilocapra americana Ord. ANTELopE.—* Becoming very scarce south of the Belle Fourche River, and entirely exterminated in the vicinity of Spring Creek.”—W. W. G. 5. Ovis cervina Desm. Mountain SHEEP.—“I was told of the presence of a small herd of Mountain Sheep in the vicin- ity of Harney Peak, in the Black Hills. In the Bad Lands they are quite common. Several were seen by our party, and their 1 Mr. Granger’s field notes are distinguished by marks of quotation and his initials (W. W. G.). 2 Described below, p. 264. 3 Cervus macrourus RAFINESQUE, Am. Month. Mag., I, Oct., 1817, p. 436. Based on the “long-tailed deer’? of Charles Le Raye’s Journal. Cf Baird (Mam. N. Am., 1857, p. 652) on the probable availability of the name Cervus macrourus Raf. for the White-tailed Deer of the Upper Missouri and Upper Platte region, as against C. Zewcurus Douglas based on a deer from the Columbia River. 264 Bulletin American Museum of Natural History. \Vol. VU, tracks could be seen at any time. ‘They live mostly in the high flat-topped buttes, where there is good grass.’’—W. W. G. 6. Lepus campestris Zach. PRairir HARE ; WHITE-TAILED Jack Rappit.—Corral Draw, May 25, 2 specimens ; Long Island, Kans., Sept. 24 and Oct. 27, 2 specimens. 7. Lepus melanotis J/earns. EasteERN BLACK-EARED Jack Rapsit.—Long Island, Kans.,. Sept. 17-Nov. 2, 8 speci- mens. 8. Lepus sylvaticus bachmani (/Vater/.). TEXAN Woop Hare.—Long Island, Kans., Sept. 15-Oct. 29, to specimens. In general features these specimens greatly resemble the form of Wood Hare of the coast region of Texas. They are, however, somewhat larger, but not otherwise sensibly different. Five adult females and three adult males measure as follows : Total length, 399 (372-432) ; tail vertebra, 55 (51-63.5) ; hind foot, 93 (89-95). 9. Lepus sylvaticus nuttalli 2ach. Nurra.i’s Woop Hare.—Corral Draw, May 23 and Aug. 22-26, 7 specimens ; Battle Creek, May 28, 1 specimen ; Cheyenne River, July 7-12, 6 specimens. Of these 14 specimens 8 are young in various stages of imma- turity. ‘They are provisionally referred to the northern interior form of the sy/vaéicus group, specimens from the type locality of nuttalli (Columbia River region) being unavailable for compari- son. 10. Lepus sylvaticus grangeri, subsp. nov. BLAcK HILLS Woop HARE. A series of 6 specimens from Hill City, in the Black Hills (Aug. ro, 11), represent a form of the sy/vaticus group very dif- ferént from that found in the bad lands and creek bottoms of the adjoining country to the eastward, the differences being shown quite as strikingly by the half-grown young of the two forms as by the adults. 1895.| Allen, Mammals from the Black Hills Region. 265 Size medium ; ears small and heavily clothed, as are also the feet ; colora- tion dark. Dorsal region dull vinaceous buff, minutely varied with black and gray, becoming purer gray posteriorly, and dull yellowish gray on the sides. Inner edge of thighs buffy; rest of lower parts pure white, with the usual pectoral collar of grayish brown. Ears small, externally dusky varied with gray, well clothed on both surfaces. Measurements.—TVotal length (average of two adults), 353; tail vertebra, 53; hind foot, 90; ear (from notch), 60. Type, No. 3933, 6 ad., Hill City, Custer Co., S. Dak., Aug. 11, 1894; W. W. Granger. Young in first pelage resemble the adults in coloration, except that the tints are duller. This is a mountain form, comparable with Z. s. pnetis of the White Mountains of Arizona (see this Bulletin, VI, 1894, p. 348), which form it strongly recalls in its dark coloration and small hairy ears. Its coloration is in strong contrast with that of the form inhabiting the adjoining open country to the eastward, the pale yellowish tints of the latter being replaced in the mountain form by pale vinaceous. 11. Erethizon epizanthus Arandt. YELLOW-HAIRED Por- CUPINE.—“‘ Not uncommon along the Cheyenne, and in the Bad Maads..——W, W.G. 12. Geomys lutescens Merriam. LuTESCENT POCKET GopHeEer.—Long Island, Phillips Co., Kans., Sept. 16—Oct. 13, 5 specimens. 13. Thomomys talpoides (Avch.). Gray Pocker GOPHER. —Corral Draw, May 23-June 14, 4 specimens; Spring Creek, July 5, 1 specimen ; Custer, July 27, 1 specimen. As shown by the above list of localities, this species is found not only in the prairie country at the base of the Black Hills, but in the small parks in the Black Hills, at an altitude of 5500 feet. 14. Perodipus richardsoni A//en. RicHarpson’s KAN- GAROO Rav.—Pendennis, Lane Co., Kans., May 8, 1 specimen ( ¢ ad.) ; Cheyenne River, Custer Co.,S. Dak., July 7-10, 3 speci- mens (1 gad. and 2 2@ im.). 266 Bulletin American Museum of Natural History. |Vol. VII, 15. Perognathus paradoxus Merriam. Larcre PocKEer Mouse.—Pendennis, Kans., May 8, 1 specimen (4 ad.) ; Corral Draw, Aug. 23-25, 8 specimens, including three quarter-grown young ; Corral Draw, Aug. 27, 1 specimen (a nursling). The young in first pelage differ greatly in coloration from the adults, the whole dorsal surface being drab-gray, with no trace of a fulvous lateral line. At the next stage, or when about one- fourth grown, they are fuliginous brown above, varied with ful- vous-tipped hairs, with, however, the fulvous lateral line well developed, but in general coloration still very unlike the adults. 16. Perognathus fasciatus Wied. Maximiiian’s PocKEr MousrE.—Cheyenne River, July 7, 1 specimen (¢@ im.); Corral Draw, Aug. 20-27, 5 specimens (4 and @ ad., and three young about one-third grown). Very young specimens show only a faint trace of the pale yellow lateral line, and they are grayer above with less olive than adults. 17. Zapus princeps 4//en. Rocky MouNTAIN JUMPING Mousr.—Two specimens (one a skull only without skin), collec- ted Aug. 6 at Corral Draw, are provisionally referred to this species. While closely agreeing with this species in coloration and in cranial characters, it differs from it in the possession of much larger ears, in this respect resembling Z. ¢/7notatus Rhoads. 18. Fiber zibethicus pallidus Mearns. Pate Muskrat.— Custer, Aug. 4-9, 7 specimens; Hill City, Aug. ro, 2 specimens. These examples seem quite as pale and as small as typical speci- mens of pallidus from Arizona. Unfortunately, however, meas- urements taken from the fresh specimens are lacking. ‘Common on nearly all of the creeks which have their origin in the Black Hills.”—W. W. G. 19. Microtus (Mynomes) longicaudus (Jd/erriam). — LONG-TAILED MEADOW MOUSE. Arvicola (Myonomes) longicaudus MERRIAM, Am. Nat. Oct. 1888, p. 935. Custer, S. Dakota. 1895. | Allen, Mammats from the Black Hills Region. 267 Three adult specimens, collected at Custer (the type locality of the species), July 25—Aug. 9. 20. Microtus (Mynomes) insperatus 4//en. BLack HiLis Merapow Mouse. Arvicola insperatus ALLEN, Bull. Am. Mus. Nat. Hist. VI, 1894, p. 347. Four specimens, Custer, July 25—Aug. 9, and one specimen, Hill City, Aug. 11, as already noted (c/. this Bulletin, 1. c.). This may be a form of MV. pennsylvanicus (=riparius auct.), but it is much paler and grayer than specimens from the Atlantic Coast region, the difference in coloration being striking. “T found these mice in the same localities as the other species [ AZ. longicaudus|. Some were caught on a hillside which was covered with aspens, and the rest along the banks of a creek.” — W. W. G. 21. Microtus (Pedomys) haydenii (4aird). Haypen’s Meapow Mouse.—One specimen, Spring Creek, June 22. (Cf. this Bulletin, VI, 1894, pp. 328-330.) 22. Evotomys gapperi brevicaudus Merriam. BiAack Hitts Rep-BacKeD Mouse.—Custer (type locality of the species), July 25—-Aug. 9, 19 specimens; Hill City, Aug. 10, 1 specimen. Eight specimens of the 20 are more or less immature. The 12 adults give the following measurements: 7 males, total length, 131 (120.6-146) ; tail vertebrae, 36 (33-39); hind foot, 19.5 (19- 20.5): 5 females, total length, 140 (130-146) ; tail vertebra, 37 (35-39.6) ; hind foot, 19.8 (19-20.5). The females thus average slightly larger than the males. Compared with £. gapperi from New Brunswick, the red of the -dorsal region is darker and the sides are much grayer, with almost none of the strong yellowish tint seen in £. gapperz. It is also somewhat smaller, 20 adults of Z. gapperi (10 24 and 10 22) from Trousers Lake, N. B., measuring as follows: Total length, 141 (130-162) ; tail vertebre, 40 (35-45); hind foot, 19.3. The corresponding averages for the 12 Black Hills specimens are 134, 36.7, and 19.6. The ears in drevicaudus, as stated by Dr. Merriam, are conspicuously larger than in gaffert. 268 Bulletin American Museum of Natural History. {Vol. VU, 23. Onychomys leucogaster (Mved7). Missouri Grass- HOPPER Mouser.—Represented by 6 specimens, all adult, collected at Pendennis, Lane Co., Kans., May 8-10. 24. Peromyscus leucopus arcticus (JZ/earns). ARcrTIC Wuire-FooTeD Mouse.—The Black Hills form of Peromyscus leucopus seems distinctly referable to avcticus. As a series the Black Hills specimens tend to a slight fulvous wash, but a large part of them fairly match a small series from Osler, Saskatchewan, received in exchange from Mr. Outram Bangs, two of which are labeled by Mr. Bangs as “almost perfect matches of the type [of arcticus, from Fort Simpson, H. B. T.] in color, length of tail, ete.” I refer to this form two series, one of 39 specimens, collected at Custer (alt. 5500 feet), July 25—Aug. 9, and another of 20 speci- mens, collected on Squaw Creek, “‘just in the edge of the pine forests of the Black Hills,’ July 20-22. The Custer series is uniformly dark, only a few specimens presenting any decided fulvous or reddish wash. The Squaw Creek series 1s similar in coloration, except that it contains one specimen (No. 9370) strongly approaching the characteristic fulvous tint of zedrascensis, to which form it should perhaps be referred. Nineteen specimens from Custer give the following measure- ments : ro males, total length, 149 (140-165); tail vertebrae, 65 (57-76); hind foot, 20.3 (19-22): 9 females, total length, 143 (128.5-162) ; tail vertebrae, 61 (50-73); hind foot, 19.3 (17.5- 20-5). 25. Peromyscus leucopus nebrascensis (J/earns'). Fut- vous WutrE-rFooTED Mousr.—Many of the specimens here referred to nebrascensis agree perfectly with the series on which nebrascensts Mearns was based (all October specimens, from the 1 Baird (Mam. N. Am., 1857, p. 462, in text, lines 5 and 6) makes the following reference to a var. nebrascensis : ** Judging from the color and the extreme shortness of tail, | am inclined to believe that Richardson’s species [M/us deucopus| is the Hesferomys sonortensis, var. nebras- censis.’ He gave no description or diagnosis, and nowhere else employed the name, which was thus a xomen nudum, till defined and duly installed by Mearns in 1890 (this Bulletin, IT, No. 4, Feb, 1890, p. 287). Apparently he intended at one time to adopt this name for the Upper Missouri specimens, but later decided to refer them to sonxorZensis, and finally based, as he states (1. c., p. 474), his description of his /esferomeys sonoriensis on specimens * from the Upper Missouri’’--in other words, the Plains region from South Dakota north to northern Montana, as shown by his table of specimens. The basis of Mearns’s ebvascensts is a series of seven specimens from the northwestern part of Custer Co., Montana, belonging to this Museum. —s ry 1895.| Aden, Mammals from the Black Hills Region. 269 plains north of the Yellowstone), but the midsummer adults are bright golden brown above, and hence much more strongly ful- vous than fall specimens of webrascens?s. The midsummer adults of the Granger collection are, however, paler, or yellower, than true fexanus. The Granger specimens referred to this form are (1) a series of 16 collected on Spring Creek, at the edge of the Bad Lands, June 19-23; (2) a series of 6 collected on Cheyenne River, bordering the Bad Lands, July 7-12; (3) a series of 34, collected at Corral Draw, in the Bad Lands, May 16-June 6; (4) a series of 11 collected at the same locality, August 19-27. The adult August specimens agree well with the type series of nebrascensis ; the May, June and July adults are many of them much brighter and more golden, while many others are not dis- tinguishable from fall specimens of the type series. Doubtless the lighter and more yellowish coloration of the May and June and early July specimens is a seasonal feature. Thirty adult specimens from Corral Draw give the following measurements: 15 males, total length, 157.5 (144-165); tail vertebrx, 66 (59-71); hind foot, 19.5 (17.5-20.5): 15 females, total length, 156.5 (144.5-173); tail vertebra, 65 (60-76); hind foot, 19.3 (17.5-20.5). There is thus practically no sexual differ- ence in size. The difference in color between the series from the arid Bad Lands (ébrascensis) and the Black Hills series (avcticus) is re- markably striking, affording an excellent illustration of the influence of environment. 26. Peromyscus leucopus texanus (/Vaserhouse). TEXAN WHITE-FOOTED Mouse.—Two specimens from Pendennis, Lane Co., Kans. (May 8-10), and one from Long Island, Phillips Co., Kans. (Oct. 6), are provisionally referred to this subspecies. 27. Neotoma campestris 4//en. PLains Woop Rav.— Pendennis, Kans., May 8 ; ro specimens, as already recorded (this Bulletin, VI, 1894, p. 322). “The ten specimens obtained were all taken in a single canen. The rocky sides of the cafon afforded excellent retreats for the 270 Bulletin American Museum of Natural History. \|Vol. VU, rats. They build large bulky nests, under shelving rocks, con- sisting of several bushels of prickly pear (cacti), “cow chips,’ sticks, and weeds. The animals were very unsuspicious and easily trapped.” —W. W. G. 28. Neotoma rupicola 4//en. Bap Lanps Rat.—Corral Draw, June 7 and Aug. 20-27 ; 35 specimens, as already recorded (this Bulletin, VI, pp. 323, 324). “This was a common species in the Bad Lands, where it lives in small caves and crevices along the ‘draws’ and cafions, and in hollow cottonwoods at the bottom of the draws. It builds small nests of cacti. It is occasionally found at the ranches along the Cheyenne Kiver) WwW. W.G. 29. Neotoma grangeri 4//en. BLack HiLts Woop Rat.— Custer, July 25-Aug. 9, 14 specimens ; Glendale, Sept. 8, 2 speci- mens. (For previous record see this Bulletin, VI, pp. 324, 325.) “Inhabits the ranches and log cabins. Nearly every deserted log cabin contained a brood of them at the time of my first visit to the Black Hills (July 24 to Aug. 11). At Glendale Mine I trapped two or three in the mill, where they had done much damage by destroying the leather lacings of the belting. A small nest is sometimes built, which is very different from that made by NV. campestris."—W. W. G. 30. Mus decumanus /7a//. Brown Rar.—Long Island, Kans., Sept. 14-16, 11 specimens. 31. Mus musculus Z7v. Houste Mousr.—Spring Creek, 2 specimens ; Squaw Creek, 1 specimen ; Glendale, 1 specimen ; Long Island, Kans., 2 specimens. 32. Sciurus ‘niger ludovicianus (Custis). WrsTERN Fox SQuIRREL.—Long Island, Kans., Oct. 7-16, 4 specimens. They are all small (probably young of the year), with the ventral sur- face wholly white, or white slightly blotched or washed with pale fulvous. They are not apparently otherwise different from aver- age Illinois specimens, which are also often white-bellied. 1895. | Allen, Mammals from the Black Hills Region. 271 33. Sciurus hudsonicus dakotensis 4//en. Briack HILis CHICKAREE.—Squaw Creek, July 21-23, 3 specimens ; Glendale, Sept. 3-5, 7 specimens, as already recorded (this Bulletin, VI, PP- 325, 326). “A common animal throughout the timber. Similar in habits to the Eastern Chickaree, from which it differs, however, slightly in its notes.” —W. W. G. 34. Tamias minimus Zach. Pate CxutpmMunK.—Corral Draw, May 16-June 7, 7 specimens; same locality, Aug. 21-27, 8 specimens, The May specimens are faded, but are not yet in molt ; the June specimens are much worn and were molting ; the August specimens have nearly all acquired the new dress, and are much more strongly colored than the May and June specimens. 35. Tamias quadrivittatus borealis Al/en. NorrHrrn CHIPMUNK.—Squaw Creek, July 24, 1 specimen ; Custer, Aug. 4, 4 specimens ; Glendale, Sept. 4, 5, 13 specimens. ‘The series is quite uniform in coloration, all of the specimens being in post- breeding dress. They are of course all from the wooded region of the Black Hills, and in coloration are in striking contrast with the series of 7. minimus from the adjoining Bad Lands. 36. Spermophilus'’ tridecemlineatus pallidus 4//en. Pave STRIPED SPERMOPHILE.—Pendennis, Kans., May 8, I specimen ; Long Island, Kans., Sept. 15-18, 3 specimens; Spring Creek, ak july 4, x1 specimen; Custer, S. Dak., July 24-29, 19 specimens. 37. Cynomys ludovicianus (Ord). Missouri PRarRiE Doc.— Corral Draw, June 25, 4 specimens; Cheyenne River, July 13, 3 specimens ; Long Island, Kans., Sept. 15, 1 specimen. The June specimens are still in the soft coat of winter, and in three of them the coarse over-hair has either been cast or has 1 Tn this volume of the Bulletin (axtea, p. 337) I hastily followed Dr. Merriam in substituting Antsonyx for Spermophilus. While Anisonyx Rafinesque is pertinent and antedates Sfervmo- philus, it is preoccupied by use ten years earlier by Latreilie (Gen. Crust. et Insect, II, 1807, p. 119) for a genus of Coleoptera—a fact that might easily have been discovered by reference to the ‘ Nomenclator Zoologicus’ of either Agassizor Scudder Dr. Merriam has recently corrected the unfortunate error (Science, N. Ser., II, No. 30, p. 107, July 26, 1895). 272 Bulletin American Museum of Natural History. |Vol. V1, worn off so that very little of it remains, leaving the general colo- ration yellowish instead of reddish. The July specimens have nearly completed the spring molt, though one of them still retains the winter coat over the posterior fifth of the body, showing that the coat is renewed in spring from the head posteriorly. ‘The September (Kansas) specimen is also in mixed coat, the summer pelage still clothing the top of the head, nape and shoulders, while the rest of the dorsal surface is clothed with the new winter coat, showing that the fall molt begins at the posterior part of the body and proceeds gradually towards the head. 38. Arctomys dacota Merriam. Buack Hitts Marmor.— Four specimens, three adult and one half grown, Custer, July 26-29. These specimens are from the type locality of the species ; they are very uniform in coloration, and agree so well with the original description as to leave nothing to be added, except to make record of the measurements for future reference. 9138 ad. Total length, 622; tail vertebrae, 179; hind foot, 84. 9139 4 ad. ¢ 635 ; <3 7 Oi si g140 ¢ ad. Ny o22— ie Wiis} p sb 84. QI41 2 juv. cy 457; My “T found the Woodchucks at an altitude of about 5000 feet in the Hills, where they were fairly common, and confined almost entirely to the rocky cliffs.” —W. W. G. 39. Castor canadensis Aw//. Braver.—No specimens were taken, but Mr. Granger contributes the following note : “Rapidly becoming exterminated. ‘There is a small colony near the mouth of Spring Creek, and two trappers took ten from Battle Creek in the winter of 1893-94.’"—W. W. G. 40. Corynorhinus townsendii (Cooper). TOwNsEND’s BaT.— Three specimens, Cheyenne River, July 8-12. Alar expanse, 308 (305-310) ; total length, 104 ; tail vertebra, 46 ; hind foot, 9.4. This seems to be a rare bat in collections. It was originally described from “Columbia River” specimens, and was recorded by Dr. H. Allen (Mon. N. Am. Bats, p. 66) in 1864 from “ Utah,” and “Upper Missouri,” the latter record being based on a single specimen collected by Dr. F. V. Hayden. In his later Monograph 1895.| Allen, Mammals from the Black Hills Region. 273 (1894, p. 60) he records only these same examples. Dobson, in 1878 (Cat. Chirop., p. 181) records a single specimen from Van- couver Island. Inthis Museum there are three specimens from Arizona (Fort Verde, Mearns; Pinal Co., Scott ; and Prescott, Keays) and one from Guadalajara, Mexico (Buller), which I pro- visionally refer to this species. 41. Adelonycteris fusca (Aeauvors). Brown Bar.—Three specimens from Corral Draw and one from Squaw Creek are very much paler than any examples of this species I have seen from other localities. ‘They seem to indicate the existence of a pale race of this species in this region of pallid forms. 42. Vespertilio ciliolabrum Merriam. LirvLe Pate Bar.— This small, pale bat is represented by 7 specimens, 6 of which are males, taken at Corral Draw, in the Bad Lands, May 16 to June 4, and one (the female) Aug. 19. The following are the average and extremes of the measurements recorded by the collector on the labels : Total length, 82 (79.5-85.5); tail vertebra, 38 (36-39.6) ; hind foot, 7.9; alar expanse, 214 (201-223). ‘This series is very uniform in coloration, the color of the upper parts being pale buffy white. 43. Scalops aquaticus argentatus (Aud. & Bach.). Si- VERY MoLe.—Long Island, Kans., Oct. 6-19, 3 specimens. ‘These are much. lighter and more ‘silvery’ than Mississippi Valley specimens. 44. Sorex forsteri “ich. Forster’s SHREw.—The single specimen of Shrew, taken at Custer, July 25, has been kindly identified by Mr. Gerrit H. Miller, Jr., as above. The collector’s measurements are: Total length, 87; tail vertebra, 36; hind foot, hi. 9ex, 2 ad. 45. Putorius longicauda Aon. Lonc-TaiLeD WEASEL.— Custer, July 29, 6 ad. Total length, 366; tail vertebre, 132; hind foot, 40. Hill City, Aug. to, 6 ad. Total length, 386; tail vertebre, 135; hind foot, 40. ‘There is an additional skull, with- out skin, from Custer. [ September, 1895.\ 18 274 Bulletin American Museum of Natural History. (Vol. V11.| 46. Lutreola vison (Sc/red.). Minx.—Long Island, Kans., Oct. 10,é ad. ‘Total length, 625 ; tail vertebra, 200; hind foot, 66. 47. Spilogale interrupta (Ra/.). BLAcK-TAILED STRIPED SKUNK.—Long Island, Kans., Oct. 10-30, 3 specimens. gi30gad. Total length, 546; tail vertebrae, 208 ; hind foot, 51. Q13I 4 juv. 7 484 ; i 181 ; oe 51. 9132 6 juv. es 433 ; A 184 ; 43. ce 48. Mephitis mesomelas Zich¢t. Texas Skunk.—Custer, Aug. 8, 2 ad.,skinand skull and an additional skull ; Long Island, Kans., Oct. 10-13, 3 specimens. These specimens are all referred to this form, to which Kansas ‘specimens seem undoubtedly to belong. The single Black Hills skin is immature, but seems to be best referred here. 49. Taxidea taxus Sc/reb. Bapcrer.— Cheyenne River, July 13, 9 juv. 50. Canis lupus nubilus (Say). Gray Wotr.—‘ Not un- common at all the localities visited. Along Cheyenne River a good many cattle and colts are annually destroyed by them. | saw three Wolves kill a calf in Corral Draw one evening.”— WeweG: 51. Canis latrans Say. Coyorre.—‘* More common than the Wolves? —W. WG. 52. Vulpes (?macrourus Zar). “There was a den of Red Foxes in one of the draws in the Bad Lands, but I was unable to obtain any of them.”—W. W. G. Bee Ie yns (sp. ?).— “Several Lynxes have been killed along Spring Creek during the last two years. They also exist in the Bad Lands and in the Black Hills in some numbers.’’—W.W.G. Probably Z. rufus and ZL. canadensis are both found in the region under consideration. Article VIII—DESCRIPTIVE CATALOGUE OF THE SPHINGIDA FOUND WITHIN FIFTY MILES OF NEW YORK CITY. By WiLLiAM BEUTENMULLER. Puiates II-VII. The present paper constitutes the second part of my work on the Lepidoptera found within a radius of fifty miles of New York City, and is the beginning of a series of similar papers on the Moths of the region. The first part of the work, on the Butter- flies, was published in the Museum Bulletin, Volume V, 1893, pp. 241-310. The main object of the work is to enable those interested in the study of our local fauna to identify their material. The de- scriptions in the following pages have been made as brief and simple as possible, and with the aid of the illustrations the species may be readily recognized. Family SPHINGID. The members of this family are commonly called Hawk-moths, on account of their powerful and rapid flight ; they are also called Hummingbird Moths, owing to their peculiar habit of hovering over flowers while drawing up nectar with their long proboscis, and while in this position they superficially resemble Hummingbirds. Some species fly during mid-day in the hottest sunshine, while others fly late in the afternoon and at night. The moths are of medium or large size, with long and narrow fore wings, with an oblique and entire outer margin or with the outer margin excavated or scolloped. The hind wings are much shorter with the outer margin entire, the anal angle usually pro- duced and the apex rounded or pointed. The head is usually clothed with smooth scales, or has a tuft between the antenne. The eyes are hemispherical, naked, and are as a rule lashed in front above. 276 Bulletin American Museum of Natural History. |Vol. VII, The proboscis is well developed in most of the species, and is nearly as long or longer than the body, and when not in use is curled up like a watch-spring between the palpi. The antenne are fusiform, ciliate in the male and simple in the female, and with the tip more or less bent into a hook. In some species the antenne are club-shaped, with a few short setee at the extreme tip. The thorax is well developed, either with the vestiture smooth, or with the posterior portion with erect scales, or with the ante- rior portion with an elevated tuft. The abdomen is long and graceful as a rule, with the segments gradually tapering, and some species are provided with a more or less complete fan-like tuft at the end of the body. The mature larvz are smooth, or sometimes more or less granu- lated over the surface. The last segment is provided with a horn, or, in absence of this, the place is marked by a tubercle or polished eye-like spot instead. The majority of the larve are provided with seven lateral oblique stripes. After they have reached matu- rity and are ready to transform they descend to the ground and burrow into the soil, where they construct a cell, in which they change to pupz. Some species, however, form their pupze on the surface of the ground, in a loose cocoon between leaves. ‘The pup are most always chestnut brown, elongate, with the tongue- case buried or detached and resembling the handle of a pitcher. Subfamily MACROGLOSSIN. Hemaris Da/man. Head small, untufted; palpi closely scaled, cone-like ; proboscis corneous, nearly as long as the body ; eyes of medium size, lashed ; antennae about two- thirds as long as the costa of the fore wings, swollen, club-shaped towards the end, which terminates in a minute and bent seta, biciliate in the males, simple in the females ; thorax smooth, closely scaled ; abdomen flattened beneath, with a broad fan-like anal tuft. Fore wings eleven-veined, transparent in the middle, the outer border somewhat rounded ; hind wings also transparent in the middle, the outer border somewhat excavated between veins 14 and 2. The members of this genus fly during the middle of the day in the hot sunshine. They hover over flowers and very much resem- ble Hummingbirds. ‘The larva construct loose cocoons on the surface of the ground, 1895.| Beutenmiiller, Hawk-moths of Vicinity of New Vork. 277 ———— Hemaris thysbe (/aiér.). Pare tle Bre. 3. Fore wings transparent, with narrow costal and inner borders, broad outer border dentate within, and base reddish brown ; the extreme base is washed with olive green scales. Hind wings also transparent, with a narrow outer border and basal half of wings reddish brown, costal border very narrow ; bor- ders of wings beneath paler than above. Head and thorax above olive green, white beneath. Abdomen above with first and second segments yellowish green, third and fourth reddish brown, and fifth and sixth olive green and reddish brown along the middle ; abdomen beneath reddish brown with small white tufts at the sides ; anal tuft reddish brown, black at the sides. Legs whitish, with the tarsi reddish brown. Expanse of wings about 2 inches=50 mm. Var. uniformis Gr. & Xob.—(Plate II, Fig. 2.) Differs from ¢hysde in having the outer border of the fore wings somewhat narrower and not dentate within. It is also less common. Size same as ¢hysée. Var. floridensis Gr. & Xob.—Similar to ¢hysée but is more robust and larger, with the borders of all the wings broader, thus making the vitreous space smaller. Expanse, 2.40 inches=65 mm. Larva.—Head pale green, with numerous minute granulations. Body pale whitish green along the dorsal region, limited by a serrated white longitudinal line along the subdorsum and running from the anterior edge of the second segment to the caudal horn ; these two lines are in close proximity on the second segment, but gradually become wider apart on the middle segment, and the space between decreases again as they meet at the sides of the caudal horn; along the dorsum are two longitudinal white stripes close together ; sides of body light green ; spiracles yellow, scarlet red in the middle ; first segment with granulations on the dorsum (representing the cervical shield) ; anterior edge of second segment with a transverse row of canary yellow elevated spots ; caudal horn blue, with black and white dots ; over the body are numerous small white dots placed regularly in transverse rows. Underside deeper green than above, sometimes partly pink; thoracic feet reddish with a black and yellow ring, extreme base yellow ; abdominal legs green with a black and yellow patch out- side, the yellow forming a stripe on the tenth and eleventh segments. Anal plates edged with yellow. Length, 1.60-2 inches=40-50 mm. Pupa.——Broadest about the middle, tapering thence to the anal extremity ; surface finely shagreened ; color brownish black ; junction of segments smooth, brown ; head-case subtriangularly produced ; tongue-case buried ; terminal spine broad at base, somewhat flattened, rugose, rounded towards the tip, with a marginal row of minute hooks on each side, and a larger double hook at the tip. Length, 1 inch=25 mm. food-plants.—Various species of Viburnum, Honeysuckle and Snowberry (Symphoricarpos). 278 Bulletin American Museum of Natural History. |Vol. VU, Very common in this vicinity, especially during the latter part of July and early in August. It is double brooded, the first brood appearing during the latter part of May and early in June. The eggs are generally deposited singly on the underside of leaves. The larva when fully grown spins a thin web-like cocoon among leaves on the ground. ‘The form wxiformis is less common than thysbe, and the form /foridensis is very rare in this district, but is common southward. The species ranges from Labrador and Canada to Florida and westward to the Mississipp1. Hemaris gracilis (Gr. & Rod.). Pras i; Pics 3: Wings transparent, with reddish brown borders, outer border broad with the inner margin straight. Thorax and first and second segments of the abdomen olive green, remaining segments reddish brown, slightly olive at the sides of the last two segments ; anal tuft reddish brown with a black tuft at each side. Thorax beneath pale yellowish white with a reddish brown stripe on each side. Abdomen beneath reddish brown with three rows of small white scale-like spots. Legs reddish brown. Expanse, 1.60 inches=40 mm. Very rare in this neighborhood, appearing in May and June and again in July and August. It is closely allied in general appearance to /Z. thysbe var. uniformis, but differs from it by its smaller size and by having a red stripe on each side of the thorax beneath and three rows of white spots on the underside of the abdomen. It is also a more graceful and slender built insect than uniforms. ‘The early stages are unknown. Hemaris diffinis (Zorsd.). PLATE LL, Fre: 4. Fore wings largely transparent, with a very narrow blackish costal border and a broader outer border of the same color, gradually narrowing as it reaches the hind angle; at the apex on the outer border is a rust colored spot; base of wings with a blackish patch elongated along inner margin. Hind wings also transparent with a very narrow outer border and a very broad inner border marked with red. Head above and thorax along the middle olive yellow, sides of thorax yellow ; the colors of the thorax are continued over the back of the basal segments of the abdomen which is black, last two segments yellowish ; anal tuft black, yellow in the middle above. Thorax beneath yellow. Legs black. Upper side of palpi black, underside yellow. Expanse of wings, 1.60 inches=40 mm. 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 279 Larva.—Head oval, green, bluish, or reddish, with fine granulations. Body bluish above, green at sides and reddish beneath, sometimes more entirely red- dish or brownish. Along the back is a median reddish shade and a whitish or yellowish subdorsal line along each side running from the second segment to the caudal horn ; laterally above the spiracles is a yellow line more or less broken ; caudal horn black or reddish ; the granulations of the cervical shield anteriorly are yellow. Length, 1.50-1.60 inches=37-40 mm. Pupa.—Similar to that of H/. ¢thysée ; in fact there is no perceptible differ- ence between them. Food-plants.—Bush Honeysuckle, Snowberry (Symphoricarpos), and Fever- wort (7riosteum per foliatum). Found during the latter part of May and early in June, and again during July and probably August. In the immediate vicinity of New York this species is very rare. It is found from Canada to Florida and westward to Missouri and Iowa, and in certain localities is rather common. Hemaris axillaris (Gr. & Rob.). Fore wings transparent with dark brown borders, the outer one broadest and dentate within, and witha dark reddish mark before the apex. Hind wings also transparent and bordered with dark brown. Head, thorax and first two segments above, olive yellow or greenish. Abdomen black, brownish on the back, the last two segments olive yellow ; anal tuft black, olive yellow centrally above. Underside of head and thorax light yellow, the latter with a black stripe on each side. Legs black. Expanse, 1.60-1.80 inches=40-45 mm. Var. marginalis Gr.—Color and size same as in 4. axillaris, but differs in having the outer border even or slightly dentate within instead of strongly dentate. Larva.—Whitish green on the dorsum, yellowish green on the sides, and dark brown on the underside. Head yellowish green, mandibles black. Cer- vical shield with yellow tubercles on each side. Spiracles black surrounded by a narrow white border. Caudal horn glossy black, yellow at the base. Thor- acic feet black. Length, 1.25-1.50 inches=31-37 mm. Pupa.—Similar to that of H. thysbe. Food-plants.—Various species of plants of the Honeysuckle family. Very rare in this vicinity, but more abundant in the Western States. It is found from New York to Texas. In fresh examples just emerged from the pupa the transparent portion of the wings is thickly powdered with black scales, which are lost in flying. 280 Bulletin American Museum of Natural History. |Vol. VU, In general appearance it resembles 4. dfinis, but the outer mar- gin is not as much rounded and the body is more elongated, and the outer border is more or less toothed inwardly, while in &finzs it is even. Synopsis of Species of Macroglossine. Hemaris. Color olive green and reddish brown ; wings transparent centrally. Discal cell crossed by a longitudinal bar of scales. Outer border of fore wings strongly dentate inwardly... ... H1. thysbe. Robust ; wings broadly bordered with reddish brown, vitreous Space: Smalls. ir. ce) ccm ewes) t Skala ete eich var. floridensis. Outer border of fore wings not dentate inwardly... ...var. wnzformis. Discal cell without the longitudinal bar of scales. Underside of thorax with reddish lateral shades.......... Hi. gracihs. Colors black and yellow. Outer border of fore wing broad and dentate inwardly. .//. axillaris. Outer border slightly or not dentate inwardly....... var. marginalis. Outer border of fore wing narrow, not dentate inwardly. ..//. diffinis. Subfamily CHGZROCAMPINE. Aellopos Hibner. Body depressed, smoothly scaled, and of almost equal width throughout ; abdomen with sides almost parallel, last segment slightly narrower, with a broad, long-haired, flat, fan-like tuft; underside of body flattened, with the vestiture from above overlapping along the sides, in form of short tufts. Head broad, prominent ; eyes not prominent, palpi pointed. The vestiture of the head, palpi and thorax are closely applied, appearing to form one piece. Legs not spinose ; middle tibize with short terminal spurs, hind tibize with two pair of spurs. Antennze of almost equal width, with a short, pointed, recurved hook at the tip. Fore wings with apex acute, costa and inner margins straight, outer margin oblique. Hind wings small, apex rounded, anal angles produced. Aellopus fadus (Cram.) PrAne inl ETGser Fore wings sooty black with an olivaceous tinge, a median, straight, narrow, whitish band from the end of the cell to the inner margin, closely followed by a narrower, second and similar band and an outer arcuate row of white spots from the costa towards the inner margin. Hind wings sooty black. Head, thorax and abdomen sooty brown with an olivaceous tinge, the third segment of the abdomen white (var. ¢¢av), or concolorous ( fadus). Expanse, 2.20- 2.30 inches=55-57 mm. 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 281 This is a Southern species, ranging northward, said to be occa- sionally found in this vicinity. The early stages are not known. It may be known by its sooty black color and white third seg- ment of the body. It flies in the day time in the hottest sunshine. Aellopus tantalus (//i/.). Similar to the preceding species, but the white band on the primaries is obso- lete and the subterminal arcuate band of whitish spots is reduced to two or three irregular spots toward the middle of the wing ; beneath the median band is entirely wanting. It is also smaller. Expanse, 1.80 inches=45 mm. It is not improbable that this insect will prove to be identical with 4. ¢antalus, when the early stages are known. it is a Southern species, and rarely occurs in this vicinity. Triptogon JIWeéneétries. Body stout, fusiform, head large and broad with prominent eyes ; palpi erect, smooth and reaching about the middle of the front ; proboscis about as long as the body ; antennz rather short, somewhat thickened towards the end, minutely ciliate in the male, simple in the female, and the tip with a short, pointed, recurved hook. Thorax smooth, with a prominent crest on anterior portion ; abdomen long, conic; anal tuft small, hardly spreading. Legs unarmed ; mid- dle and hind tibiz spurred. Fore wings shorter than the body, inserted before the middle of the thorax ; apex obtuse and excavated beneath to vein 5, where it is again excavated to the anal angle, which is decidedly produced; inner margin excavate for some distance before the apex. Hind wings with apex rounded, somewhat excavated before the anal angle, which is produced into a broad angle. Triptogon lugubris (Zzvm.). Wings and body chocolate brown; fore wings with a darker shade outwardly, and before the middle is an oblique, narrow transverse line; across the wings are traces of transverse lines, but they are very indistinct. Abdomen with two rows of dark spots along the back, which are absent in some individuals. Underside uniform chocolate brown with traces of transverse lines. Expanse, 2-3 inches=50-75 mm. Larva.—Head dark green, with a yellow frontal band. Body pale green with dark green dorsal dashes and a dark green subdorsal line, bordered beneath with whitish ; along each side are nine pale yellow oblique bands ; spiracles reddish. Length, 2.40 inches=60 mm. Food-p lants.— Grape and Virginia Creeper A mpelopsis). 282 Bulletin American Museum of Natural History. [Vol. VU, This Southern species occasionally occurs in this neighborhood, but very rarely. It is common in the Southern States, Mexico, and the West Indies. Amphion /iéner. Head small, not sunken into the thorax; palpi forming a point in front of the head ; eyes lashed above; tongue almost as long as the body ; antenne fusiform extending to a little beyond the middle of the costa, ciliate beneath in the male, simple in the female, hooked at the end. Thorax well developed and much broader than the head, scales rather closely applied. Abdomen narrowing suddenly to the tip which is provided with a flat fan-like tuft, with the sides rounded and the middle prominently pointed ; posterior edge of segments at the extreme sides with short tufts of scales. Anterior and middle tibie with a few short spines near the tip; middle and hind tibia with small spurs. Fore wings as long as the body, outer margin excavated below the apex, and again above the hind angle, which is prominently produced ; inner margin concave before the angle. Hind wings with apex rounded, . outer margin excavated before the anal angle which is obtusely produced. ; a Amphion nessus (C7.). PLATE II, Fic. 5. Fore wings rich dark brown, crossed by a darker velvety brown band which is divided at the costa, and with shades of the same color before the outer border ; on the costa before the apex is a reddish brown patch, and across the outer third is a narrow buff colored transverse streak, which is.sometimes quite obsolete. Hind wings rich, deep brown, with a median reddish brown band. Wings beneath rusty yellow, brown at the outer borders ; across the middle of the wings are two very narrow ferrugineous transverse lines. Head, thorax and abdomen rich, deep brown, the latter with a narrow canary yellow transverse band between the fourth and fifth segments. Underside of head and thorax rust colored with a yellowish line on each side ; abdomen darker, with two small white tufts on the posterior edge of the segments. Expanse, 1.80-2.50 inches=45-62 mm, Larva.—Uniform chocolate brown, checkered with black markings and dotted with dark amber, especially along the back, and there are stripes of the same ,color along the sides ; caudal horn reddish. Sometimes the larva is of a bluish green color, with the stripes yellow and seven oblique lateral stripes of the same color. Length, 2.50-3 inches=62-75 mm. Food-plants.—Epilobium, Ampelopsis, and Grape. Not common in this vicinity. It is found late in May and early in June, and flies in the hottest sunshine and also in the evening. Found from Canada to Florida, and westward to lowa. 1895.] Beutenmiller, Hawk-moths of Vicinity of New York. 282 Sphecodina Blanchard. Body stout, depressed, with the sides almost parallel. Head well developed, with the scales forming a low ridge between the antenne ; palpi closely scaled and forming a blunt point in front of the head ; eyes moderate, distinctly lashed above. Antenne of almost equal width, ciliate in the male, simple in the female, tip gradually tapering into a bent hook. Thorax well developed, rather broader than long, scales smooth in front, rather loosely tufted transversely behind ; abdomen as broad as the thorax, last segment slightly tapering, underside somewhat flattened, anal tuft of male long and flat, with a short point in the middle; in the female the tuft is long and cylindrical, with a short lateral bunch of hairs ; posterior edges of last segments with raised scales, and laterally with bunches of scales, forming dentations. Legs unarmed, middle and hind tibiz spurred. Fore wings almost parallel, longer and narrower than the body, apex produced and excavate below and also above the hind angle, the excavate portion with a minute dentation ; hind angle produced and excavated on the inner margin. Hind wings dentate along the outer margin, anal angle somewhat produced. Sphecodina abbotii (Swazns.). . PLATE II, Fic. 6. Fore wings chocolate brown with a narrow, black oblique line running from the basal third of the inner margin and terminating on the costa at about the middle of the wing ; beyond this the wings are paler and several dark streaks run from the inner margin towards the outer border and are lost about the middle wing, above which the lines are strongly dentate, but indistinct. The terminal space is variable dark brown. Hind wings bright yellow at the base, outer border deep blackish brown, containing several pale streaks at the anal angle. Underside of fore wings brownish yellow, with a narrow, strongly dentate black line across the outer third, beyond which the wings are chocolate brown. Hind wings beneath bright yellow at base, costal region scaled with brownish, outer border chocolate; through the middle of the wings are two dentate lines. Head and thorax above chocolate brown with a bluish iridescence; across the thorax are two narrow black transverse lines ; abdomen blackish at base, paler and with more or less iridescence across the middle, darker at tip. Expanse, 2.20-2.80 inches=55-70 mm. Larva.—Chocolate brown, with very narrow transverse lines ; a dark dorsal line, and one of the same color along the subdorsum and sides ; anal segment provided with a large polished spot instead of a horn. Head dark with a lighter broad band on each side. Sometimes the larva is marked with numerous pea-green patches, oval on the back and irregularly triangular on the sides, with an interrupted subdorsal chocolate-colored line. Head brown with a light green band on each side. Length, 2.20 inches=55 mm. 284 Bulletin American Museum of Natural History. |Vol. VII, Pupa.—Deep chestnut brown, paler between the segments. Head-case broad and rounded ; tongue-case concealed and level with the breast ; segments deeply punctured, smooth between the segments; last segment with a rugose, wedge-shaped point. Length, about 1.40 inches=35 mm. Food-plants,—Grape and Virginia Creeper (4 pelopsis). Common in this neighborhood. The moth appears in May and early in June, and again during the latter part of July and early in August. When at rest the larva does not assume the attitude of holding up the head common to the larve of Sphingide, but stretches out at full length. If disturbed it throws its head from side to side, hereby producing a creaking noise. It is found from Canada and eastern United States westward to Iowa. Deidamia Clemens. Head small, with a prominent tuft between the antenna ; eyes small and lashed ; palpi rather short; antennae of almost equal width and tapering at the apex which is bent at the tip, but not into a recurved hook, biciliate in the male, simple in the female ; thorax stout, vestiture forming a distinct dorsal ridge. Abdomen conical, anal tuft in the male small, flattened laterally, and forming a rounded bunch in the middle ; in the female the tuft is composed only of a bunch of short hairs. Wings longer than the body, and very similar in outline to Amphion, only somewhat less oblique. Deidamia inscripta (//ar/7s). PLATE, Ll re. 7: Fore wings ashen gray with minute brown scales ; before the middle is a transverse brown band incurved on the inner margin, and before this band are some indistinct transverse lines; median space gray; across the outer third is a broad brown band, angulated outwardly above the middle and shaded out- wardly with brown ; in the pale apical region is a rich brown spot, and a short dentate apical streak. All these markings are ill defined, somewhat confluent, and more or less suffused with brown, Hind wings reddish brown with a some- what darker terminal band. Underside of fore wings dull fawn color, with an irregular ferrugineous transverse shade outwardly, beyond which the space is dark with a distinct white spot in the apical region ; hind wings ashen brown with indistinct transverse lines. Head and thorax ashen brown, the latter with three more or less distinct transverse whitish lines edged with brown. Abdomen with a row of small brown spots along each side of the back. Expanse, 1.80- 2.40 inches=45-60 mm. ee 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 285 Larva.—Body pale green with a yellow subdorsal line ending at the base of the caudal horn, which is whitish at the tip. The segments are also trans- versely marked with fine black lines. Length, 2 inches=50 mm. Pupa.—Dark pitchy brown, mottled with testaceous on the wing-cases, tho- rax and head-case, and also somewhat on the segments ; head-case with a pointed tubercle, and also one on each eye-case, which are surrounded with a sharp ridge ; tongue-case concealed, keel-shaped ; the leg and antennz-cases are also out- lined with sharp ridges ; anterior part of thorax with a transverse mark outlined and divided in the middle by a ridge; segment deeply punctured, last one provided with a sharp spine. Length, .80-1.20 inches=20-30 mm, Not common in this vicinity. The moth makes its appearance during the latter part of May and the first days in June, and is probably double brooded. ‘The larva is fully grown about the last of June or early in July. It feeds on the Grape and Virginia Creeper (Ampelopsis). Deilephila Ochsenheimer. Body stout fusiform ; head of moderate size, not sunken into the thorax, smoothly scaled ; eyes rather large ; tongue nearly as long as the body ; antennx gradually thickening toward the tip, which has a minute, pointed hook ; thorax smooth ; abdomen smooth, segments gradually decreasing in size ; tip pointed, and provided with a bunch of long hairs, forming the anal tuft. Fore wings with apex and hind angle acute, outer margin oblique, entire, inner margin very slightly sinuate. Hind wings with outer margin entire, and a slight projection before the rounded anal angle. Deilephila lineata (/ad;.). PEATE Wh biG. 2: Fore wings pale olive brown with a broad oblique buff band running from near the base of the inner margin to the tip of the apex, where the band terminates in a point. The veins are heavily marked with white to the outer space, which is lilac gray. Hind wings black, with a broad, pink median band ; outer border narrow, pinkish, with the fringes white. Head, thorax and abdomen olive brown ; thorax with three parallel white stripes on each side; abdomen with a row of large black and white spots on each side, and along the middle is a narrow broken white line with a small black spot on each side at the posterior end of the segments. Underside lilac gray with the oblique buff band of the fore wings partly repeated ; the pink band of the hind wings is buff color; the wings are also minutely dotted with brown. Expanse, 3-3.60 inches=75-90 mm. 286 Bulletin American Museum of Natural History. \Vol. VU, Larva.—Yellowish green, with a prominent subdorsal row of elliptical spots, each spot consisting of two black curved lines, inclosing superiorly a crimson space and inferiorly a pale yellow line ; the whole row of spots is connected by a pale yellow stripe edged above with black ; sometimes these eye-like spots are disconnected, or the larva is black with a yellow line along the back anda series of pale yellow spots and darker yellow dots along the sides. This dark form is subject to variation, some specimens entirely lacking the line along the back, and having the spots of different shape. Length, 2.50-3 inches= 62-75 mm. Pupa.—Light brown, head-case prominent, showing the palpi, rugose, as is also the thorax ; segments punctured ; tongue-case not apparent. Length, 1.60-1.80 inches=40-45 mm. Food-plants.—Purslane (Portelaca), Buckwheat, Turnip, Watermelon, Chick- weed (Stellavia), Dock (Rumex), Evening Primrose, Apple, Plum, Currant, Grape and Gooseberry. This species is found in the United States, Canada, and Cuba. In this vicinity it is common everywhere. It flies early in the evening and often in bright daylight. The larve are most com- monly found in fields feeding on purslane, which seems to be their favorite food plant. The insect is double brooded, the first brood appearing during June and July, and the second during the latter part of August and early in September. Deilephila galii, var. intermedia Avrdy. BEATE, Lil IRiGss3 Fore wings olive brown with a buff-colored oblique band, running from the hind margin near the base to the apex; the upper edge of the band is indented, and the lower edge somewhat curved ; the outer border of the wings is lilac gray, base black. Hind wings with a broad pink central band, followed by a narrow black band; terminal border lilac gray, base black. Thorax olive with a white stripe on each side, running along the head. Body olive with a row of minute white dots along the middle; first and second segments marked with black on the sides ; the second, fourth, and following segments with white. Underside of thorax dull yellowish brown ; wings buff color with the marking from above somewhat reproduced. Expanse, 2.65-3 inches=65—-75 mm. Larva.—Dark green with nine yellow spots encircled with black on each side ; spiracles yellow with a black ring ; caudal horn red ; head and thoracic feet blue gray. Sometimes the larva is olive green with a bright yellow dorsal line and spots on each side ; or is blackish gray with a red dorsal line and two rows of yellow spots on each side. Length, 3 inches=75 mm, i 1895.| Beutenmiiller, Hawk-moths of Vicinityof New York. 287 Pupa.—Pale brown mottled with black in its impressed portions ; head-case projecting, corrugated ; tongue-case concealed, corrugated ; wing-cases corru- gated ; segments punctate ; terminal spine tapering, bifid at tip. Length, 1.80 inches=45 mm. Food-plants.—Epilobium, Purslane (Portulaca), Evening Primrose and Apple. Not common in this neighborhood. Found during June and again in August. It occurs from Canada to Georgia and westward to California; also in Europe. The median pink band of the hind wings in the European form (ga/?) is much paler than in intermedia. Theretra iibner. Body long and graceful; head of medium size and smoothly scaled ; eyes large, hemispherical ; proboscis almost as long as the body. Antenne minute- ly ciliate in the male, simple in the female, and of almost equal width throughout, with a small bent hook at the tip ; thorax smooth ; abdomen very long, and tapering to a point, vestiture very smooth and closely applied, and a narrow brush of hairs at the tip. Fore wings long and narrow, apex sharply pointed, outer margin very oblique, inner margin sinuate ; Hind wings narrow, apex pointed, sharply produced before the anal angle. Theretra tersa (Ziwz.). PEATE ILI, Fig. 4: Fore wings ochreous brown, paler at the base, which has a slight purplish reflection ; from the apex to the middle of the inner margin is a series of from seven to nine diverging, oblique light brown lines; the outer ones are almost parallel with the outer border, and the inner ones extend nearly to the base of the wing. Hind wings smoky black with a series of large wedge-shaped, yellow subterminal spots. Head and thorax fawn color, with a roseate lateral stripe ; abdomen rusty brown above with indistinct fawn-colored stripes; sides rusty yellow ; underside paler. Expanse, 2.25-3 inches = 56-75 mm. Larva.—Pale green, with fine longitudinal irrorations ; along the subdorsum is a rather broad white band, running from the fourth segment to the caudal horn ; on the band is placed, on each segment, a round black ring, the one on the fourth segment with a black eye-like spot in the centre. Caudal horn reddish, tip black. Length, 2.40 inches=60 mm. Food-plants.—Bouvardia, Buttonweed | Spermacoce glabra), Manettia bicolor. 288 Bulletin American Museum of Natural History. |Vol. VU, ~ (Quite rare in this vicinity, but common in the Southern States, West Indies, Central and South America. It ranges northwardly as far as Canada. It is usually found in flower gardens. ‘The moth may be easily recognized by its graceful form. It varies somewhat in ground color from light to dark ochreous brown. Argeus /Hiibner. Form robust ; head large and prominent; eyes large, not lashed; thorax stout, well advanced in front of the base of the fore wings ; abdomen long and robust. Fore wings shorter than the body, outer margin obliquely rounded, inner margin sinuate ; hind wings entire with anal angle produced. Argeus labrusce (Z/n7z.). BEATE Wve) 1G: (6; Fore wings green, with a large V-shaped deeper green space; beyond this are two somewhat deeper green transverse lines ; a series of small black sub- terminal spots, and a rather large rounded brown patch about the middle of the wings. Hind wings blue, with a median black band, in which is a blue spot, and marked with red towards the inner margin ; in front of the outer border is another black band ; outer portion of wings buff colored. Head, thorax, and abdomen green; abdomen with a series of white lateral spots. Expanse, 4.25-4.75=106-I119 mm. A South American species, occurring northward to Canada. In the north it is an occasional visitor, and is very rarely taken. Pholus Aitiner. Body robust ; head large, smoothly scaled ; eyes large, not lashed ; antennce ciliate in the male, simple in the female, hooked at the tip; thorax stout and smooth ; abdomen stout, smooth and gradually tapering. Tibiz not spinose, middle pair with two unequal terminal spurs, hind tibize with two pairs. Fore wings much longer than the body and broad, outer margin rounded and some- what excavated below the apex or obliquely rounded, inner margin strongly sinuate. Hind wings broad, apex pointed, outer margin somewhat excavate before the anal angle. Pholus pandorus /iibver. PAgiO Ue EGS. Fore wings pale olive green, marked with patches and shades of rich olive green ; on the inner margin near the base, is a long olive green patch running 1895.] Beutenmiiller, Hawk-moths of Vicinity of NewYork. 289 to about the middle of the wing, and separated by a pink shade from the triangular patch within the hind angle; a similar patch rests on the costa a little before the apex, there is also a dark olive shade from the costa extending outwardly ; across the wing are several narrow, wavy, indistinct lines ; discal spot consisting of two or three small blackish spots at the end of the cell. Hind wings pale greenish at base, with a black patch at the middle of the inner margin, and a broad black subterminal band running from the costa to the middle of the wing, where it breaks up into spots and lines on a roseate ground ; outer margin olive green. Head and thorax with a blackish line along the middle, thorax at sides with an olive green triangular patch. Under- side pale olive, with two narrow transverse lines on each wing. Expanse, 3.75-4.50 inches=9g5-11I2 mm. Larva.—Pale green above, darker at sides, or reddish brown, lighter dorsally ; at the sides from the fifth to tenth segments each with an oval cream-colored spot in which are the spiracles. Last segment with a black, polished, eye-like spot instead of a caudal horn; anterior segments with numerous, minute black dots. Head small and rounded. Length, 3 inches=75 mm. Pupa.—Elongate, chestnut brown. Head-case prolonged, subtriangular laterally, slightly corrugated ; wing-cases smooth ; tongue-case concealed, and extending slightly beyond the wing-cases ; segment thickly punctured ; middle portion of pupa thicker than either extremity; terminal spine rather long, pointed and minutely bifid. Length, 2.50 inches=62 mm. Food-plants.—Grape and Virginia Creeper (Ampelopsis). Rather common in this neighborhood, in gardens and vineyards. It is double brooded ; the first brood appearing during June and early in July, and the secondin August. It is found in the United States east of the Great Plains and also in Canada. Pholus achemon (Drury). PLATE III, Fic. 6. Fore wings pale chocolate brown with a pinkish tinge, with darker shades and several wavy transverse lines ; on the inner margin at the middle is a large, deep velvety brown quadrate patch, and a small triangular patch of the same color before the hind angle, and a larger one on the costa immediately before the apex. Hind wings pink with a light chocolate-brown outer border, containing a row of deep brown spots, which are not clearly defined before the middle of their course. Head, thorax and abdomen same color as the upper wings, the thorax with a deep triangular patch on each side. Underside of fore wings roseate with a pale chocolate-brown outer border and with two parallel, deeper brown transverse lines ; hind wings roseate brown, powdery, with two transverse lines. Expanse, 3.75-4.25 inches=g95—106 mm. [ September, 1895.} ao 290 Bulletin American Museum of Natural History. |Vol. VII, Larva.—Varying from green to reddish brown or pale straw-color, darker along the sides ; an interrupted brown line runs along the middle of the back, and an unbroken one extends along each side, beneath which are six cream- colored oblique spots, one on each segment ; on the last segment is a polished black, flat tubercle, instead of a caudal horn. Over the body are sprinkled numerous minute dark dots. Length, 3 inches=75 mm. Pupa.—Very similar to that of P. pandorus, but less elongated, and the head- case less prolonged and pointed ; the last segments are broader and the terminal spine shorter and obtuse, instead of pointed. Length, 2-2.25 inches=50-56 mm, food-plants.—Grape and Virginia Creeper (A mpelopsis). This species is double brooded, the first brood appearing in June and July, and the second in August. It is somewhat com- mon in this vicinity, and is found throughout the United States from the Atlantic to the Pacific, and also in Canada. The larva, in shape and size, is the same as that of P. pandorus, but differs in shape of the spots along each side, which are elongated instead of oval as in pandorus. Pholus vitis (Z7mz.). BX TUNE MMe, 775 Fore wings deep olive green, with a pale flesh-colored band extending from the middle of the base to the apex and crossed by a similar band which runs from the outer fourth of the costa to the middle of the inner margin ; the costa to the transverse band is chocolate brown, as is also the outer border ; between the oblique transverse band and the outer border the veins are pale flesh- colored ; at the basal third of the wing is a narrow line running from the band to the inner margin. Hind wings pale greenish inwardly, with a broad pink outer border interrupted by an-olive green space before the angle; within the outer border is a broad black fascia terminating in two narrow lines ; below the disc are two large black spots, separated by a few pink scales. Head, thorax and abdomen flesh-colored, head and thorax with a median olive green line, and an elongate olive green patch on the patagia ; abdomen with a blackish patch on each side of the base, and on each side of the back a broad olive green stripe. Underside of wings pinkish flesh color, roseate at the inner margin of the hind wings. Expanse, 3-4.25 inches=75-106 mm. Larva.—Head claret red with two black stripes ; body yellowish, with narrow transverse black lines, junctions of segments claret red ; along the sides are a series of elongate, oblique, oval, whitish patches ; beneath greenish with black lines. Length, 3 inches=75 mm. 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 291 Pupa.—Shape similar to that of 7. pandorus, but differs in having only the anterior parts of the segments, as is also the prolonged head-case ; the last seg- ment is more rounded, and the terminal spine is somewhat curved and polished at the base. Length, 2.20 inches=55 mm. Food-plant,—Grape. This is a southern species very rarely taken in this vicinity. It has been recorded from South America, Central America, Cuba, Texas, Florida, and along the Atlantic coast to Massachusetts. The larva is very different from its congeners. Pholus linnei (G. & 2.). Closely allied to P. vitis, but differs in having the outer border of the fore wings rounded instead of being straight as in vi¢/s. The markings of the fore wings are darker, and the hind wings are greenish at base, with the outer border grayish brown, instead of pink, and at the anal angle is a pink patch and a black patch within. Expanse, 3.25-4.5 inches=81-1.12 mm. Inhabits South and Central America, Cuba, and the Southern States, and is said to be found northward as far as Massachusetts ; if so it may possibly be found in this vicinity. Everyx Méntétries. Head small, vestiture forming a central ridge or tuft between the antennz ; eyes moderate and slightly lashed ; proboscis about half as long as the body ; antennz ciliate in the male, simple in the female, with a large hook at the tip. Tibia not spinose, except in £. charilus, which has the anterior and middle pair minutely spinose. Thorax prominent, smooth ; abdomen untufted, smooth and tapering. Fore wings as long as the body, rather broad, apex subfalcate, outer margin more or less excavate, inner margin sinuate. Hind wings excavate before the anal angle, apex rounded. Everyx cheerilus (Cramer). PLATE II, Fic. 8. Fore wing rusty brown with slight purplish reflections ; basal half grayish brown with two curved transverse lines ; outer part of wings rusty brown crossed by several more or less distinct zig-zag lines ; terminal space same color as the base of the wings. Hind wings uniform rusty red with a very narrow fringe, edged with white. Head, thorax and body rusty brown, patagia edged with gray. Underside of wings pale rusty brown with two faint transverse lines on each ; outer border of fore wings same as above. Expanse, 2.25- 3 inches=56- 75 mm. 292 Bulletin American Museum of Natural History. |Vol. VII, Larva.—Body pale green, with a darker dorsal line ; second segment yellow- ish green with numerous irrorations. Spiracles orange, white above and below, those on second segment are orange and yellow above and below; along each side of the 5th, 6th, 7th and 8th segments is a white oblique band, and on the posterior segments the bands are connected with a continuous line to the base of the caudal horn, which is bluish at the base, green at the tip and white in the middle. Sometimes the color of the larva varies from pinkish to brown or even leaden brown. Length, 2.20 inches=55 mm. Pupa.—Purplish brown, with a pink tint over the whole surface, and slightly mottled with black. Wing-cases mottled with black, spaces between the seg- ments blackish brown. ‘longue-case concealed. Length, 1.40 inches=35 mm. Food-plants.—Sheep-berry (Viburnum Jlentago), Arrow-wood (Viburnum dentatum), also other species of Viburnum, Sour-gum (Vyssa), and Azalea. This is a rather common species, and is found in open woods. It may be easily recognized by its rusty brown color and purplish- gray shades on the fore wings. It is double brooded. The larva spins a rude cocoon, amongst leaves on the surface of the ground. Found from Canada to Georgia, and westward to Iowa. Everyx myron (Cramer). PLATE II, Fic. 9. Fore wings olive gray, varying to purplish gray, with olive green, oblique, transverse bands and shades, which are more or less distinct. The band across the basal third is continuous, while the band across the outer third is nearly always more or less broken in the middle by the pale ground color. ‘The outer part of wing is shaded with olive green at the apex and inner angle, leaving the rest of the terminal space olive or violet gray. Hind wings rusty brown with a darker, more or less distinct and complete terminal band. This band is very often reduced to an olive gray patch on the anal angle. Head, thorax and abdomen olive green or gray. Underside pale rust red or grayish, with two narrow transverse lines. Expanse, 1.80-2.50 inches=45-62 mm. Var. cnotus //ziéz.—In this variety the fore wings are uniform brown, without traces of the olive transverse bands and shades. Larva.—Body green sprinkled with yellow dots ; along the middle of the back is a row of yellow patches, each containing a spot varying from red to pale lilac ; along the sides, from the head to the caudal horn, is a white stripe with a dark green margin, and below this are seven oblique lateral stripes. The caudal horn varies from red to bluish, granulated with black, and is sometimes yellow behind and at the tip. The larva is sometimes green, cream color, purplish brown, deep brown, or leaden brown. In some the yellow patches along the back are almost wanting. Length, about 2 inches=50 mm, wets. c tes ; om KW . 7 = ie Ww = 1895.| Beutenmiiller, Hawk-moths of Vicinity of NewYork. 2923 Pupa.—Shape like that of £. cherilus. Pale brown ; wing-cases sprinkled with black dots; junctions of segments dark brown. Tongue-case concealed. Length, about 1.20-1.40 inches=30-35 mm. Food-plants.—Grape and Virginia Creeper (Ampelopsis). Rather common in gardens about grape vines. It may be readily known by its olive gray color with olive green markings. The larva in shape is like that of /. chwrilus, but differs in having the lateral oblique stripes interrupted by the lateral white line, which forms a straight edge at the junction of the oblique lines. It also differs in having a row of sub-oval dorsal spots, which are absent in &. cherilus. When fully grown it spins a loose cocoon amongst leaves on the ground. It is double brooded, the first brood appearing in June and July and the second in August. It is found from Canada to Georgia, and westward to Missouri and Iowa. The variety cvotus is rare, but is the common and prevail- ing form in the Southern States. Everyx versicolor (Harris). PLATE TT Bie. 10: Fore wings green shading into lighter green or yellowish green, with a num- ber of transverse whitish or pinkish lines. Several more or less distinct curved lines run from the costa before the middle to the base of the wing, and beyond the middle, across the outer fourth, are also three more or less distinct transverse lines and a curved subterminal white line beginning at the apex and running irregularly down to the anal angle ; near the apex it is crossed by two white dashes ; sometimes this line is almost absent. Hind wings rust brown, the margin grayish. Head, thorax and abdomen green tinged with yellow ; a narrow white line extends.over the top of the head to the end of the abdomen ; along the sides of head and thorax also a whitish line. Underside of wings marked with green, yellow and white, sometimes with reddish on the fore wings. Expanse, 2.75-2.90 inches=68—-72 mm. Larva.—HUead and first four segments yellowish green ; rest of body pea green, with a white line along each side from the mouth to the base of the caudal horn ; this line is composed of several lines, as follows: a subdorsal line, extending from each side of the mouth back to the rear of the fourth segment of the body ; a similar line runs obliquely from the lower part of the fourth segment, just under the stigmatal point, upwards and backwards to the rear of the fifth segment, meeting it just below the dorsal line. This is followed by five other parallel lines, each beginning and ending one segment further back, except the last, which extends across the last three segments up to the base of the caudal horn. There are also faint indications of other lines at the lower part of the 294 Bulletin American Museum of Natural History. {Vol. V1, tenth and eleventh segments ; body also covered with white specks. Spiracles red, with yellow at each end. Caudal horn black, red on the sides. Sometimes the ground color is pinkish brown instead of green, and the markings are then pinkish white. Length, 2.50-3 inches=62-75 mm. Pupa.—Dirty brown, with chocolate brown spots, almost covering the wing- cases and anterior parts ; eyes and spiracles black, as also between the seg- ments ; tongue-case concealed. Food-plants.—Buttonbush (Cephalanthus occidentalis) and Swamp-loose- strife (Vesea verticillata). Quite rare and local in this vicinity. It is double brooded, appearing in June and early in July and again in August. The moth may be easily known by its bright green shades on the fore wings, with the more or less distinct whitish transverse lines. The larva is very quiet in its habits, never leaving a stem of the food-plant so long as a leaf remains. In eating, when- fully grown, it hangs from the mid-rib of the leaf, and eats usually from the extreme end, finishing a section across the leaf as it goes. It generally eats the mid-rib and petiole down to the woody stem. Synopsis of Species of Cherocampine. Aellopos. Abdomen with a broad fan-like anal tuft. Wings entire. Sooty black ; abdomen with the third segment white. Fore wings with two transverse bands beneath the cell and an arcuate TOW OL SUbtenmInally SpOtSacr 6)).. sect) eeu eee A. fadus, Fore wings with the bands obsolete, and the subterminal row of Spots’ reducediitostwo)or three Spots. 4. sini ee 4. tantalus. Triptogon. Fore wings excavate, below the apex and above the hind angle. Chocolate brown with darker shades outwardly on the fore wings. Thorax with a prominent tuft on the fore part......... T. lububris. Amphion. Fore wings excavate outwardly ; rich dark brown with darker markings. Abdomen with a transverse, canary yellow line between the fourth andififthisegments “oan 8.0) Sac peepee er ee eee A. nessus. Sphecodina. Fore wings excavate outwardly ; last segments of abdomen with dentate lateral tufts. Dark brown, fore wings with oblique streaks ; hind wings yellow at . DaSGzAE Mahi: Seb bneahlc > « Naahinan ele teats eeetn eae S. abbotii, 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 295 Deidamia. Fore wings excavate outwardly, head with a prominent tuft. Ashen gray with brown markings ; hind wings reddish brown, D. inscripla. Deilephila. Wings entire, not excavate ; fore wings with a buff-colored oblique band from the base to the apex. Veins of fore wings and thorax lined with white......... .... D. lineata. Veins and thorax not lined with white........... D. galii var. intermedia. Theretra. Body long and graceful, wings pointed, entire. Pale ochreous ; fore wings with paler oblique stripes; hind wings black with a subterminal row of yellow wedge-shaped spots, 7”. lersa. Argeus. Outer margin of fore wings not excavate beneath the apex. Fore wings wholly green, with a darker V-shaped space. ind wines: blue, black, red) and’ buif.-2...-....-...-.4 4. labrusce. Pholus. Large species ; wings broad; fore wings slightly excavate beneath the apex. Fore wings pale olive green, with dark green shades and patches. Hind wings pale green with a black patch and subterminal band, P. pandorus. Fore wings pale chocolate brown, with rich, dark brown patches. Hind wings pink, outwardly pale brown........... ....P. achemon. Fore wings deep olive, with an oblique flesh-colored band from the middle of the base to the apex, and an oblique transverse band of the same color. Hind wings pale green at base, pink outwardly, and with a black Sporandishori band sors .crecr | -lenicic wei ers nite P. vitis. Hind wings pale greenish at base, black outwardly; anal angle witlagpimke PAtGan aeeiqena as ste alae: ser aes cies es - P. linnet. Everyx. Fore wings with apex subfalcate, entire. Olive gray, with olive green markings, hind wings red brown... £. myron. Rusty brown ; basal half of fore wings pale grayish brown...Z. charilus. Light and dark green, with white transverse lines on fore wings ; hind wings ferrugineous ; from head to end of body a white line, E. versicolor. Subfamily SPHINGIN/#. Dilophonota Aurmeister. Body long ; fore wings as long as the body ; head large, smoothly scaled ; eyes large ; tongue about half as long as the body ; antennz minutely ciliate in the male, simple in the female, tip with a short hook ; thorax smooth, with a short, divided rest on the middle of the anterior portion ; abdomen smooth, 296 Bulletin American Museum of Natural History. |Vol. VXI, slender. Legs unarmed; middle and hind tibie spurred. Fore wings with the outer margin oblique and slightly scolloped between the nervules. Hind wings with apex acute, anal angle somewhat produced. Dilophonota ello (Zizz.). PLATE IV, FIGs. I AND 2. Fore wings ashen gray, with an indistinct, dentate line running from the outer fourth of the costa to the middle of the hind margin, and an outer row of small dark spots. Often there is a brown shade from the middle of the base to the apex, and above and below this shade the wings are also marked with brown, forming no regular pattern. Hind wings ferrugineous, terminal border blackish, not reaching the anal angle, and grayish before the angle. Head and thorax ashen gray or marked with brown. Abdomen gray with five large, transverse, oblong spots on each side, with the space on the back forming a line. Underside of wings ferrugineous, dusky outwardly ; body light gray. Expanse, 3-4 inches=75-I100 mm. Larva.—Body green ; head with a dark brown line on each side in front, thence running over the top of the head and along the subdorsum of the body and converging at the base of the caudal horn ; these lines are bordered with yellow ; on the fourth segment is a large, round, velvety black spot bordered with yellow, and outside on either side is a shade of deep reddish brown ; on the middle of the back from the head to the end of third segment is a fine dark brown line ; abdominal feet with a velvety black patch externally ; caudal horn short and blunt. The body is also sprinkled over the surface with minute dark brown and yellow dots. Sometimes the larva is reddish brown, with the lines less distinct than in the green variety. Length, 3-3.50 inches=75-87 mm. Pupa.—Pitchy black, smooth, and very shining wing-cases and breast with longitudinal ochreous lines ; thorax and head-case also with ochreous lines ; abdominal segments ochreous with short black transverse lines and dots, last three segments pitchy ; anal spine smooth, compressed, and in form of a trian- gular tooth. Head-case prominent and smooth. Length, 2-2.25 inches= 50-56 mm. Food-plant.—E up horbia. Very rare in this vicinity but common in the South. It is found from Brazil northward to Canada. It varies from an almost uniform gray to a form with distinct brown shades. Phlegethontius /iibner. Head very large and prominent; eyes large ; proboscis much longer than the body ; antennz rather strongly biciliate in the male and simple in the 1895.] Beutenmiiller, Hawk-moths of Vicinity of New York. 297 female ; thorax robust, well advanced in front of the base of the fore wings, and with short erect tufts posteriorly ; abdomen tapering, untufted. Fore wings well developed, somewhat longer than the body, outer border obliquely rounded and entire; apex acute; hind wings with outer margin very slightly scolloped. Phlegethontius quinquemaculatus (//aw.). PLATE IV, Fic. 3- Fore wings ashen gray, shaded more or less with blackish beyond the middle and toward the apex; across the wings from the costal third is a series of three narrow black lines running outwardly, then obliquely backwards to the inner margin near the base ; across the outer fourth are three much angulated, parallel lines running to the middle of the inner margin; beyond these lines isa less angu- lated, darker and more distinct line, and a short apical streak of the same color. All the lines are more or less ill-defined. Hind wings pale gray, ashen gray outwardly, and are crossed by two parallel dentate lines through the middle and with a rather broad subterminal black band, widest at the costa. Head and thorax ashen gray, the latter marked with several short transverse lines in front, one along the edge, and another through the middle of the patagia ; across the hind part of the thorax is a broad black band, with, in front of it, one or two spots composed of biuish tufts. Abdomen gray with a row of large, conspicu- ous, orange colored spots surrounded with black along each side and edged on the anterior parts above and below with white. Underside of wings uniform ashen gray, with two transverse bands on each, the outer one on the fore wings and those on the hind wings toothed. Expanse, 3-4.50 inches=75-112 mm. Larva.—Body dull green with yellowish white, oblique lateral stripes, or dull sea green with ocellated spots anteriorly with the oblique lateral bands and stigmatal stripe flesh color ; head black with an elongated flesh colored triangu- lar patch ; thoracic feet, exterior of prolegs, and anal plates black. Underside paler than above. Sometimes the body is very dark brown sprinkled with yellow dots ; and the lateral oblique bands and the stigmatal stripe are yellow, as is also the border of the caudal shield; head with triangular spot drab; caudal horn black, spiny. Length, 3-4 inches. Pupa.—Chestnut brown ; tongue-case detached, very long, prominent, and strongly curved, like the handle of a pitcher, with the bulbous end touching the body beyond the middle of the wing-cases. Length, about 2.50 inches= 62 mm. ; tongue-case, 30 mm. Food-plants.—Tobacco, Tomato, Jamestown-weed (Datura), Matrimony Vine (Lyctum vulgare), and Ground Cherry (Physalis viscosa). Not common, and double brooded in this vicinity. The first brood appears in June, and the second in August. It may be known by its ashen-gray color and the five orange spots on each 298 Bulletin American Museum of Natural History. (Vol. V1I, side of the abdomen. Found throughout the United States and Canada. The pupa is a well-known object of interest, and from its long arched tongue-case may be readily distinguished. Phlegethontius carolina (Zzwv.). PLATE IV, Fic. 4. Fore wings dark brownish gray, with the transverse lines near the base and across the outer fourth very similar to those of /P. cefews. The outer dark transverse line is scalloped, while in ce/ezs it is slightly curved, and beyond this is a subterminal angulated whitish line. Discal spot small, white. Hind wings gray, with three transverse black bands, outer portion dark gray. Head and thorax brownish gray with yellow scales, the latter with indistinct black lines and black across the posterior part. Abdomen wood brown or gray with a row of large, deep orange spots along each side, which decrease in size towards the end of the abdomen and are surrounded with black. Underside of wings gray with transverse bands. Expanse, 3-5 inches=75~-125 mmi. Larva.—Green, paler above, with seven oblique white bands, bordered above on each side with bluish or dark brown, last segment edged with white ; caudal horn reddish, white at the base of sides, or wholly black. Over the body are also scattered fine, short, transverse lines. Length, 2.30-3.50 inches =70-80 mm. Pupa.—Chestnut brown, and similar in shape to that of P. celews, but less swollen at the middle, the detached tongue-case is shorter and much less arched, and does not quite extend half-way to the end of the wing-cases. Length, about 2.20 inches=55 mm. ; tongue-case, 23 mm. Fooa-plants.—Tomato, Tobacco, and Jamestown-weed (Datura). Closely allied to P. cedeus, but is a much darker insect, the yel- low spots on the abdomen are much larger, and the central bands on the hind wings are straight instead of toothed. It is found in the United States from the Atlantic to the Pacific, in Canada and the West Indies. In this vicinity it is sometimes quite common and double brooded. Phlegethontius cingulatus (/ad;.). PLATE EV, PIG 5. Fore wings dark gray, sometimes mottled with brown; darker than in /?, celeus, and lighter than in P. carolina, ‘The transverse lines are almost like 1895.| Beutenmiller, Hawk-moths of Vicinity of New Vork.. 299 those of P. carolina. Below the cell, between the veins, are two black streaks. Hind wings rose colored at the base, gray outwardly, and crossed by three black bands. Head and thorax dark gray, the latter with black lines at sides and in front; abdomen dark gray with five bright rose colored spots on each side, decreasing in size towards the posterior end of the body, and separated by black bands. Underside of wings dark gray; hind wings white at base along the inner margin. Expanse, 3.75-4.50 inches=9g4-112 mm. Larva.—Dark green with seven oblique black bands along each side, which terminate on the back in two longitudinal stripes of the same color; on the dorsum of the third and fourth segments are two black spots, four very small ones on the tenth, and two very large ones placed laterally at the incisure of the first and second segments. Head green with black stripes. Caudal horn yellow or ferrugineous, tip black; anal shield orange yellow. Length, 3-4 inches= 75-100 mm. Variety A.—Clear green, with the oblique lateral bands entirely white, and the two dorsal stripes replaced by two rows of black points. Variety B.—Dull green with six longitudinal rows of blackish or brownish spots, and the head and horn ferrugineous. Variety C.—Dead leaf brown on the back, white on the sides, and flesh colored beneath ; seven oblique lateral stripes of deeper brown and a lateral stripe of straw color, which is continuous on the first three segments, and inter- rupted after the fourth segments at the middle of each. Head pale fawn color with black lines ; caudal horn black ; shield orange. Variety D.—Brown with four longitudinal lines of dirty white on the first three segments, two dorsally and two laterally. Variety &.—Karthen brown with the back and oblique bands of a deeper brown. Pupa. Brown, with the tongue-case detached, not reaching the middle of the wing-cases, bent downward and backward for about half its length; the turned portion resting on the breast. Length, 2.50 inches=62 mm. Appears in June and again during the latter part of August and in September. It may be known from the other members of the genus by the rose red spots on the abdomen. The larva is very variable, and besides the varieties described above, intermediate ones are met with. It hides itself at the base of the plant under leaves, but may be discovered from its large excrements. The insect is found from Canada to Brazil, and to the extreme west of our continent ; also in the Hawaiian Islands. ; 300 Bulletin American Museum of Natural History. |Vol. VU, Phlegethontius rusticus (/aér.). PLATE TV, Bic: 6: Fore wings sooty brown, with white transverse wavy lines across the basal and outer third, and subterminal white markings of no regular pattern ; some- times the wings are more or less rust brown. Hind wings sooty brown with a whitish band near the base and two more or less distinct, black central bands, followed by whitish shades. Fringes black, cut with white. Head, thorax and abdomen sooty black, or marked with rust brown, thorax marked with white; abdomen with three large orange spots on each side. Underside of fore wings paler than above with transverse lines ; hind wings whitish or grayish with two dentate central bands; body white. Expanse, 3.50-5.60 inches= 87-140 mm. Larva.—Head and body dark green, yellow on the dorsum ; along each side are seven oblique blue bands edged with purple, and beneath these is a white longitudinal band, edged with yellow on the lower part. Caudal horn yellow, with reddish tubercles. Length, 4 inches=100 mm. Pupa.—Chestnut brown, and similar in shape to that of P. carolina, but larger and more robust, with the detached tongue-case more curved. Length, about 2.50-2.80 inches=62-70 mm. Food-plants.—Chionanthus, Privet, and Lilac. Found from New York southward into South America. In this vicinity it is very rare. It may be known by its large size, sooty- brown color, the white wavy lines, and by having three yellow spots on each side of the abdomen. Sphinx ZLinneus. Head moderate ; proboscis as long as the body; eyes small, usually distinctly lashed ; palpi curved upward and projecting beyond the head ; antennz with the tip more or less bent ; thorax well developed, untufted, metathorax with erect hairs. Legs more or less spinose; middle and hind tibiz spurred. Fore wings with very acute apex and the outer margins very oblique, entire. Hind wings narrow, apex distinct. Sphinx drupiferarum 4. & 5S. PRATE VV. RiGser: Fore wings smoky black, broadly grayish or whitish along the costal region from the base of the wings to nearly the apex ; the outer margin is also grayish, containing a whitish line which is limited inwardly by a wavy black line ; between the veins in the dark portion of the wings are several black dashes. Hind 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 3,01 wings black, base whitish, and also a central whitish band ; terminal border dirty grayish brown. Thorax black, sides grayish, as is also the head. Abdo- men brownish, with a fine black line along the middle ; sides black, with a row of four or five large white spots. Underside: fore wings smoky, terminal border grayish brown ; hind wings grayish at base and a central band and outer border of the same color. Expanse, 3-4 inches=75-100 mm. Larva.—Bright apple green; head with a_dark brown stripe on each side; along each side of the body are seven oblique white stripes bordered with purple on the upper side; spiracles orange ; caudal horn dark brown, yellow at the base of the sides. Length, about 3.50 inches=87 mm. Pupa.—Dark chestnut brown ; tongue-case detached, short and straight, and not resting on the breast. Length, about 2 inches=50 mm; tongue-case, .25 inch=6 mm. Food-plants.—Apple, Plum, Hackberry (Ce/¢zs). Not common. Double brooded in this vicinity, appearing in June and again early in August. The species is recognizable by the smoky black fore wings and the whitish costal space. Found from Canada to Florida and westward. Sphinx kalmiz 4.& S. PLATE V, Fic. 2. Fore wings pale chestnut brown, with lighter and darker streak-like shadings ; before and parallel to the outer border is a pale brownish white transverse line, limited inwardly with black ; fringes alternately brown and whitish. Hind wings brownish white with a central and subterminal blackish band. Head and thorax chestnut brown on top, patagia edged with black; sides of head and thorax pale whitish brown. Abdomen chestnut brown along the back with a narrow black line; sides black, with a row of large whitish spots. Underside of wings chestnut brown, with a terminal dark brown shade, and a paler central band across the hind wings. Expanse, 3-4.50 inches= 75-112 mm. Larva.—Body apple green, paler above and dark at the sides, with seven oblique lateral stripes, which are whitish along the middle, bordered with blue- black anteriorly and with yellow posteriorly. Caudal horn light blue, with black tubercles. Spiracles pale orange, thoracic feet black, whitish at their bases ; abdominal legs with two black spots externally, and separated by yellow. Length, about 3 inches=75 mm. Pupa.—Deep chestnut brown, with the detached tongue-case short and straight, and is similar in shape to the pupa of S. drufiferarum. Length, about 2 inches=50 mm. Food-plants.—Laurel, Lilac, Privet, Chionanthus, and Ash. 302 Bulletin American Museum of Natural History. |Vol. VU, Not common. It may be known by its chestnut brown color, with lighter and darker streaks. The larva is quite conspicuous, with the lateral oblique stripes very broad. It is double brooded, the first brood appearing in June and the second late in July and early in August. Found from Canada to Georgia, and westward to Missouri. Sphinx lucitiosa Clemens. PEATE VeceIGs3: Fore wings rusty brown with the costa and outer margin sooty brown ; most of the veins are finely marked with blackish. The band on the outer margin gradually narrows as it nearly reaches the apex ; before this band, from the inner margin, is a light brownish oblique wavy streak. Discal dot small, whitish. Hind wings ochre yellow with traces of a dark central band and a broad black outer border. Thorax and top of head sooty black ; sides of head and thorax pale brownish gray. Abdomen dull ochre yellow, with a narrow black dorsal stripe, a black band along each side, broken by whitish spots on the edges of the segments. Underside of wings pale ochreous yellow, with a smoky brown outer border. Expanse, 2.50-3 inches=62-75 mm. Larva.—Head pale green with a yellow line on each side, indistinctly edged above with black ; mouth parts black ; body green; the first three segments and the lower half of all the others covered with small white dots, each dot encircled with black ; on each side are seven oblique stripes, white, with pinkish lilac above. Thoracic legs white with red tips; abdominal legs green with a faint purplish tinge ; caudal horn rather short, green with a black stripe on each side; spiracles red. Length, 3.50 inches=87 mm. Pupa.—Bright mahogany brown, with a short detached tongue-case. Length, 1.59 inches=37 mm ; tongue-case, .125 inch=3 mm. Food-plant.—Willow. Very rare in this vicinity. It may be known by its brown color and ochre yellow hind wings. Sphinx gordius Cramer. PLATE V, FIG. 4. Fore wings gray, more or less clouded with sooty brown ; veins finely marked with black, with a few dashes of the same color between them and in the cell ; outer border of wing sooty black, the band gradually narrowing as it reaches the apex. Discal spot white. In some individuals there are traces of a few 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 303 transverse dark bands, one across the basal fourth, one across the middle, and another a little beyond. Hind wings dirty white, with a central black band and a broad black outer border. Head and vertex of head sooty black, sides gray- ish. Abdomen grayish, with a black dorsal line ; sides black and white in form of bands. Underside of fore wings smoky ; hind wings marked as above, but paler. Expanse, 3-3.60 inches=75-90 mm. Larva.—Bright apple green, with a yellow and brownish stripe on each side of the head ; along the sides of the body are seven short, oblique stripes, which are white and margined above with carmine. Caudal horn black, green on top and beneath. Length, about 2.50 inches=62 mm. Pupa.—Deep brown, with a very short detached tongue-case. Length, 1.40 inches=35 mm. Food-plants.—Apple, Pear, Ash and Wax-Myrtle (AZyrica). This species is allied to .S. Zucztiosa, but may be separated by its gray color and differently marked hind wings, which are ochre yellow in 5S. Zucitiosa. It is not common in this vicinity, and is double brooded. It ranges from Canada to Georgia, and west- ward to the Mississippi, and probably further westward. Sphinx chersis (/iézer). PEATE OV BbiGanb: Fore wings light ashy gray with a small bunch of blackish scales at the base of the inner margin ; between the veins, from beneath the cell to the apex, is a short black dash between each, the last one almost uniting with a short black apical streak ; before the outer margin is a narrow transverse black and whitish line. Hind wings whitish with a black central band and a terminal band. Head and thorax light gray ; patagia lined inwardly with black, and a tuft of the same color on each side of the posterior edge of the thorax. Abdomen gray, with a central black line ; sides black and broken by white cross-stripes. Underside of wings gray with an ill-defined subterminal band on the fore wings and a pale dentate, median band on the hind wings. Expanse, 4-5 inches=100-125 mm. Larva.—Pale apple green ; dorsal region whitish, and with seven oblique yellow stripes along each side. Head with a yellow stripe on each side ; caudal horn pale bluish ; thoracic feet pink. Length, about 3 inches=75 mm. Pupa.—Deep chestnut brown, with a short detached tongue-case. Length, 2.50 inches. Food-plants.—Lilac, Privet and Ash. 304 Bulletin American Museum of Natural History. [Vol. VU, Double brooded in this vicinity, appearing in May-June and again late in July. It may be known by its uniform gray color, very oblique outer margin and pointed apex. Found from Canada to Georgia, and westward to California. Sphinx canadensis Zoisd. PLATE VI, FIc. 1. Fore wings light brownish gray, streaked with black between the veins, and a terminal black line edged with whitish and followed by another within. Hind wings pale grayish at the extreme base, followed by a broad median and a terminal band, leaving the space between them very narrow. Head and thorax grayish, the latter brownish gray, with the patagia edged inwardly with black. Abdomen gray black on each side, broken by white on the edges of the segments. Expanse, 3.25-3.60 inches=81-g0 mm. It is possible that this rare species may occur in this vicinity. It is found in Newfoundland, Canada, Maine, northern New York and Ohio. The early stages are unknown, Sphinx eremitus Drury. PVATE Vile SRiGe 2: Fore wings brownish ash color, clouded with darker brown, with a rather heavy short black dash between each vein from beneath the cell to the apex. Discal spot white. From near the base, on the inner margin, are two short, parallel, oblique, black streaks and across the outer portion of the wing is an indistinct transverse, curved band. Hind wings with a black patch at the base and a broad median and terminal border of the same color; spaces between these and the basal spot dirty white. Head and thorax brownish ash, the latter with a black line through the middle of the patagia. Abdomen brownish ash, a median black line and the side alternately black and dirty white. Underside of fore wings grayish brown with faint indications of three transverse bands ; hind wings dirty whitish, with the central and terminal band brownish. Ex- panse, 2.50-3.15 inches=63-79 mm. Larva.—Head small, brown with a lateral white stripe. Abdominal segments reddish brown with many tan-colored or whitish ocellated spots. Second segment light brown above, olive at the sides ; collar light brown outlined with black. The second segment is subtriangularly produced, with the apex rounded, pointing forward, and extending over the head when the larva is at rest ; it is olive brown at sides with a velvety brown spot. Third and fourth segments also 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 305 olive brown with a velvety brown spot on top. Along the sides of the body are seven whitish oblique stripes, bordered with brown posteriorly. Caudal horn brown. Length, 2.25-2.75 inches=56-68 mm. Pupa.—Chestnut brown ; head-case subtriangular ; tongue-case exerted, dark brown, nearly straight and slightly raised from the breast by its bulbous end. Length, 1.60 inches=65 mm. Food-plants.—Spear Mint (AZentha), Wild Bergamot (AZonarda), Salvia, etc. Very rare in this vicinity. It may be recognized by its brown color and the black streaks on the fore wings. Double brooded. The larva may be known by the triangular protuberance on the second segment, differing in this respect from all our Sphinges, Sphinx plebeius /aé;. PEATE Vi-° Fic: 3. Fore wings gray, streaked with black between the veins from the base of the inner margin, thence obliquely to the apex. Discal spot white. Across the outer fourth are traces of transverse lines, but they are very indistinct or absent; before the outer margin are some indistinct shadings of whitish. Hind wings smoky brown with traces of a lighter shade across the middle and base ; fringes white, cut with smoky brown. Head and thorax gray, with a black line around the anterior portion and continued along the middle of the patagia. Abdomen gray, with a narrow black dorsal stripe and a broad black stripe along each side containing a row of grayish spots. Underside of wings fuscous. Expanse, 2.65-3 inches—66-75 mm. Larva.—Body green, paler above, with seven oblique lateral stripes along each side, yellow, edged with black anteriorly. Caudal horn blue, with small black tubercles. Sometimes the larva is olive pink with numerous flesh-colored dots, with the oblique bands olive, and a shade of the same color along the subdorsum ; sides of body olivaceous Length, 2.60 inches=65 mm. Pupa.—Chestnut brown, with the detached tongue-case very straight and closely applied to the breast ; and at its base it is very slightly curved and reaches to about the middle of the wing-cases. Length, 1.40 inches=35 mm. food-plant.—Trumpet-vine ( Tecoma). This species is not rare in this vicinity ; especially found in flower gardens about the trumpet-vines. It is double brooded, appearing in June and again latter in July and early in August. Found from Canada to Florida and westward to the Mississippi. [ September, 1895. 20 306 Bulletin American Museum of Natural History. (Vol. VII, Chlenogramma Smith. Form robust ; head rather large with a small tuft between the antennz ; eyes moderate, not lashed; antennz fusiform, minutely biciliate in the male, simple in the female, slightly curved at the tip ; thorax stout, somewhat produced be- fore the base of the fore wings; abdomen long, pointed, with a row of loose tufts along the back. Fore wings long, much longer than the body, and rather broad ; outer margin slightly rounded, oblique, and somewhat excavated above the hind angle. Hind wings rounded. Chlenogramma jasminearum (Boisduva/). PLATE VJ, Fic. 5.- Fore wings pale gray, finely mixed with brown and blackish scales ; across the basal third are two indistinct wavy lines, and three similar transverse lines beyond the middle, and beyond these is another line less distinct and inter- rupted ; from the basal line on the costa is a conspicuous blackish shade running obliquely to the middle of the outer margin ; discal spot small, whitish, with a yellowish brown blotch beyond ; fringes white and black. Hind wings brown- ish black with traces of a very indistinct paler central band. Head, thorax and abdomen gray ; thorax with a blackish transverse line in front extending through the middle of the patagia, hind part with two black spots and two spot- like bands along each side. Underside of wings uniform fuscous, with a slightly darker central band. Expanse, 3.20-4.25 inches=80-105 mm. Larva.—Pale green, lighter dorsally ; body with six oblique lateral white stripes, and a seventh red stripe which extends to the green caudal horn ; thor- acic feet pink ; spiracles white encircled with black, Length, 3 inches=75 mm. Pupa.—Dark brown, with a very short cylindrical tongue-case, bulbous at the end, and applied to the breast. Length, about 2.50 inches=62 mm. Food-plants.—Various species of Ash (/raxtnus). This species may be easily known by its gray color and distinct oblique black shade across the fore wings. In this vicinity it is quite rare, and is probably double brooded. It is found from Canada to Georgia. The larva and pupa are imperfectly known. Ceratomia Harris. Head rather small, with a light tuft between the antennz ; eyes small, not prominent ; antennz biciliate in the male, simple in the female ; proboscis reaching the end of the thorax ; thorax short, somewhat advanced in front of the base of the fore wings ; abdomen untufted, tapering. Spurs of middle and 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 307 posterior tibiz small. Fore tibiz with short stout terminal spinules ; anterior tarsi with three claw-like spines on the first joint and one on the second. Fore wings large, much longer than the body, outer margin oblique or slightly rounded. Hind wings with margins entire, and anal angle slightly produced ; apex obtusely rounded. Ceratomia amyntor (//bner). PEATE. VL. RIG. «6: Fore wings dark coffee brown ; costal region and outer margin clay brown ; outer portion of costa mixed with gray ; along the middle portion of the wings between and parallel with the veins, below the median vein and beyond the cell, are several heavy black dashes; at the basal portion of the wings are traces of several wavy transverse lines; across the outer fourth are several angulated and wavy transverse lines, curved outwardly. Discal spot distinct, whitish. Hind wings clay brown, with a central and subterminal ill-defined band. Head above, collar and front part of thorax whitish; patagia coffee brown with black lines ; central portion of thorax clay brown ; abdomen clay brown, with a black dorsal stripe and two lateral stripes. Underside of wings brown with grayish scales, fore wings with traces of the outer transverse lines from above repeated ; hind wings with a double transverse zig-zag line. Fringes above and below brown, interrupted with whitish spots. Expanse, 3-4.25 inches= 75-106 mm. Larva.—Pale green, sometimes reddish brown; head and body strongly granulated, with a dorsal row of short fleshy teeth, tipped with white or pink, these teeth extending from the fourth segment to the caudal horn ; on each of the third and fourth segments is a pair of short, straight tubercles, covered with short spines ; along the sides are seven oblique whitish stripes composed of granulations. Caudal horn green, thicky covered with short spines. Length, 2.75-3.25 inches=68-81 mm. Pupa.—Chestnut brown ; head small, rounded, slightly depressed ; eye-cases margined inferiorly by an impressed line ; tongue-case buried ; antennz-cases granulated ; first and second segments with a slightly elevated median line. Abdominal segments punctulate, wrinkled posteriorly ; terminal spine rugose, pointed, minutely bifid. Length, 1.50-2 inches=37-50 mm. Food-plants.—Various species of Elm, Linden and White Birch. Rather common in this vicinity in June and July. It may be readily known by its coffee brown and clay brown colors and by the black dashes on the fore wings. The larva may be known by the two fleshy horns on each of the third and fourth segments. Its favorite food is elm. The insect is found from Canada to Virginia, westward to Missouri and Iowa. 308 Bulletin American Museum of Natural History. {[Vol. VU, Ceratomia undulosa (/Va/ker). PrADE VAR EGs 775 Fore wings gray mixed with light brownish scales ; across the basal portion are three angulated transverse black lines, furthest apart on the costa and com- ing closer together as they reach the inner margin ; these lines are more or less distinct ; from the middle to the outer fourth are four transverse curved lines, the middle two are toothed and the space between them is gray; from below and beyond the cell, between the veins, are three black dashes and a wavy apical streak ; discal spot white narrowly bordered with black ; fringes alter- nately brown and white. Hind wings grayish brown, with three ill-defined blackish bands. Head grayish, thorax gray mixed with yellow; through the middle of the patagia is a black line connected with the one across the anterior portion of the thorax ; across the hind part of the thorax is also a black line edged with yellow and white ; abdomen grayish brown with a black dorsal line and two rows of large black spots along each side. Wings beneath almost uniform grayish brown, hind wings somewhat paler, both wings with two trans- verse bands ; fringes same as above. Expanse, 2.80-4.60 inches=70-115 mm. Larva.—Pale green, smooth ; on each side of the body are seven oblique yellowish white stripes ; spiracles pink ; caudal horn pink ; head green with a whitish band on each side. Length, 3 inches=75 mm. Pupa.—Dark brown ; head-case rugose, rounded, somewhat prominent, and a little compressed laterally ; eye-cases rough, slightly prominent, with a crescent-shaped mark before the middle ; tongue-case concealed ; thorax sha- greened ; wing-cases very slightly wrinkled ; abdominal segment deeply punc- tured, smooth on the junctions ; anal spine short, rugose, pointed. Length, about 1.75 inches=43 mm. Food-plants.—Lilac, Ash and Privet (Ligzstvum). Rather common and double brooded in this vicinity, the first brood appearing in June and the second in August. The ground color of the insect varies somewhat from light to dark gray, and the lines are more or less heavily marked. It is found from Canada to Carolina, and westward to Iowa. Dolba Watlker. Head rather small, roughly scaled, inclining to form a tuft between the anten- nx ; palpi roughly scaled ; proboscis longer than the body ; eyes moderate, lashed ; antennce minutely biciliate in the male, simple in the female, hooked at the end ; thorax stout, as broad as long. ‘Tibiz: not spinulose, middle with one 1895.| Beutenmiller, Hawk-moths of Vicinity of New York. 309 air sterior with two pairs of unequal spurs. Fore wings as long as the bod ’ db £ y; rather broad ; outer margin entire, very slightly excavated above the hind angle. Hind wings broad, apex well rounded, entire, and slightly excavated before the anal angle. Dolba hylzus (Drury). PLATE-VI, Fic. 4. Fore wings dark rusty brown with three transverse black lines across the basal third, the inner two diffused with white ; across the outer fourth are four dentate black lines, the outer ones with more or less white between them ; terminal space shaded with white, forming no definite pattern. Discal spot small, white. Fringes rusty brown, cut with white. Hind wings smoky brown with two parallel whitish lines across the middle. Fringes white, cut with brown. Head and thorax rusty brown, the latter white at the extreme sides, with black and white on top. Abdomen rusty brown, a dusky broken central line, and a row of small white dots on each side ; laterally the abdomen is black and broken with white on the edges of the segments. Underside of wings smoky brown, with traces of transverse lines across the outer fourth of the fore wings. Hind wings whitish and crossed by several dentate lines, followed by whitish shades. Expanse, 2.20-2.60 inches=50-65 mm. Larva.—Pea green, with seven oblique lateral whitish bands edged above with pink ; caudal horn purple, and a pale blue stripe on each side of the head. Length, 2.33 inches=58 mm. Pupa.—Chestnut brown; tongue-case concealed, straight and closely ap- plied to the breast. Length, 1.60 inches=40 mm. ; tongue-case, .50 inch= 12 mm. Food-plant.—Ink-berry (/lex glabra). Not common in this vicinity. It may be easily known by its small size and rusty brown color, with the transverse black lines and white shades. In general appearance it resembles a miniature Phlegcthontius rusticus. It is found from Canada to Florida, and westward to Lowa. Lapara Walker. Head small, retracted, the scales forming a tuft or ridge between the anten- nz ; palpi short, slender, porrect not pointed upwards, and much shorter in the female than in the male ; eyes of medium size, slightly lashed ; proboscis very short ; antennz biciliate in the male, simple in the female, tip slightly bent ; thorax scarcely advanced in front of the wings, short, stout ; abdomen un- tufted. Fore wings entire, with outer margin obliquely rounded. Hind wings with apex rounded, outer margin with hind angle somewhat produced. 310 Bulletin American Museum of Natural History. |Vol. V1, Lapara coniferarum (4.& S.). ]ereyNanioy WWI, Tie, Ti Fore wings leaden gray, with a darker, dentate line across the outer fourth, from the costa to the inner margin ; between the veins beneath the cell are two blackish dashes, the lower one being the shortest. Hind wings grayish brown. Head and thorax leaden gray ; abdomen brownish gray. Underside of wings uniform brownish gray. Expanse, 2-2.50 inches=50-62 mm. Larva.—VYellowish green with three longitudinal, equidistant, white stripes along each side. Head conical, flattened in front, yellowish green with a black line along each side, uniting on the summit. Stigmatal spaces marked with red. Caudal horn wanting. Sometimes the larva is checkered with light and dark-gray squares. Length, 2.25-3 inches=56-75 mm. Pupa.—Cylindrical, pitchy black. Head, thorax, and anterior margin of wing-covers rugosely punctate, as are also the fore margins of the segments. The four posterior segments are rugosely punctate nearly over their entire sur- face. Anal spine pointed. Tongue-case concealed. Length, 32 mm. Food-plants.—Various species of Pines. This species is somewhat variable. The fore wings are some- times almost uniformly leaden gray without the two dark dashes, or have only one dash. The insect is quite rare in this ne hood. Found from Canada to Florida. Lapara bombycoides var. harrisii (C/em.). PLATE VII, Fic. 2. Fore wings gray, with a pair of dentate lines across the outer fourth from the costa to the inner margin, with the space between light gray, as is also the outer portion of the wings ; before the middle and across the basal fifth are two transverse lines, angulated outwardly above the middle ; beneath the cell are two conspicuous black dashes. The spaces between the lines are more or less shaded with light gray. Hind wings uniform brownish gray ; also the head, thorax, abdomen and tip of the patagia are light gray, Underside of wings uniform gray or brown, sometimes with a dark line across each, Ex- panse, 2-2.25 inches =50-56 mm. Larva.—The ornamentation consists of alternate green and white longitudi- nal stripes. Dorsal stripe green, spotted with red. Head red in front, with a white or pinkish white border. Collar and legs green; prolegs and last segment bordered with red. Caudal horn absent. Length, 2-2.50 inches= 50-62 mm. 1895.| Beutenmiiller, Hawk-moths of Vicinity of New Vork. 311 Pupa.—Chestnut brown, with a rough, not produced head-case. Tongue- case buried. Posterior segment tapering. Terminal spine black, contracted at base, minutely bifid at tip. Length, 1-1.10 inches=25-27 mm. Food-plants.—Various species of Pines. Rare in this neighborhood, and is probably double brooded. Found from Canada to Florida, and westward to the Mississippi. It may be recognized by its small size, gray color, and transverse, dentate and angulated lines. Synopsis of Species of Sphingine. Dilophonota. Thorax anteriorly with a short crest divided in the middle; eyes large. Fore wings light gray or streaked with fuscous; hind wings FUSE Te Wilh. DidckOULemDOLER. 2...) se eset sees ws D. ello. Phiegethontius. Thorax not crested ; head and eyes very large; abdomen with yellow or pink spots along each side. Fore wings light ashen and dark gray, with black lines and streaks; hind wings with two central dentate lines, P. qguinguemaculatus. Fore wings dark brownish gray, markings similar to guingze- maculatus, hind wings with central lines not dentate, P. carolina. Fore wings dark gray, lines and streaks black; hind wings at base and spots on abdomen pink........ ........ P. cingulatus, Fore wings sooty or rust brown, with black transverse lines and Shadediwithinwiter sy eee a= SCs .2 ais 5. 2's peers CUSLICUS: Sphinx. Head moderate, eyes small ; abdomen black at’sides, with spots. Thorax dark smoky brown, pale grayish at sides. Fore wings smoky black ; costal region whitish. ... : S. drupiferarum. Fore wings entirely sooty black marked with gray, costal region concolorous: “Discalspotwhite::i— -.2%.f.- 22.5.5. S. gordius. Fore wings rust brown ; discal spot wanting; hind wings ochre yellow moutersbordersblacksme 42m os acidle oe. oleae S. lucitiosa, Thorax chestnut brown, sides grayish. Fore wings light and dark chestnut brown, in form of streaks, S. almie. Thorax dark grayish, sides whitish; patagia with a black line inwardly. Fore wings gray, with black dashes between the veins... S. canadensts. Thorax entirely gray ; patagia with a black line inwardly. Fore wings light gray with black dashes..... ............ S. chersis. Thorax ashen brown with black through the patagia ; sides pale. Fore wings ashen brown with black dashes.......... ..-S. evemitus. Thorax ashen gray, white at sides ; patagia with a broad black line through the middle. Fore wings gray with black streaks ; discal spot white ; hind Vala EO Ha Skic eur Su ane ee EO ae eee aoe S. plebetus. Dolba. Small species: Head roughly scaled, with a tuft between the antennz. Fore wings sooty brown or rust brown, with transverse dentate black lines, shaded with white........ ...........- D hyleus. 312 Bulletin American Museum of Natural History. |Vol. VII, Chlenogramma. Large species : Eyes large ; abdomen with bunches of raised scales along the dorsum ; legs not spinose. Fore wings gray mixed with brown, and a black shade from the middle of the costa to the middle of the outer margin, C. jasminearum., Ceratomia. Large species : Eyes small ; abdomen without dorsal tufts ; legs spinose. Fore wings coffee brown, paler along costal and outer region, and with black streaks between the veins... .. ...C. amyntor. Fore wings gray, with transverse black lines............ C. undulosa. Lapara. Small species : Head small, retracted ; palpi very small, not ascending. Fore wings leaden gray, with a transverse black line beyond the middle, sometimes with two short black dashes beneath the 2 | eee See cia coho 0 A ans rier o Me L. coniferarum. Fore wings gray scaled with white; two angulated transverse lines before the middle, and two dentate ones beyond, L. bombycotdes var. harrisii. - Subfamily SMERINTHINZ:. Amorpha /iibner. Large species ; head small, sunken, with a small median crest; palpi very small, rudimentary ; antenne strongly biciliate in the male, simple in the female. Thorax stout, not tufted ; abdomen plump, last segment blunt. Middle and hind tibize with a single pair of short terminal spines; anterior with a stout curved spine at the inner side of the tip. Tarsi finely spinose. Fore wings broad, much longer than the body, with the outer border regularly scolloped. Hind wings also scolloped. Amorpha modesta (//arris). Prare Villy Pic. 7. Fore wings light mouse gray at the base ; across the middle is a very broad olive gray band; outer portion of wings olivaceous with lighter transverse shades. Hind wings gray, shaded more or less with claret red through the middle, and at the anal angle a bluish gray patch. Head, thorax and abdomen mouse gray, witha bluish tint. The wings have also a decided bluish reflection. Underside of wings olive gray with pale gray transverse shade ; fore wings more or less claret red towards the base ; sometimes the wings are almost uni- form gray. Expanse, 3.50-5.50 inches=88-138 mm. Larva.—Light green coarsely granulated with white points ; along each side with seven oblique yellow bands, and on each side a yellow subdorsal line ; caudal horn rudimentary; prolegs red; spiracles brown; head green, trian- gular, granulated. Length, 3 inches=75 mm. 1895.| Beutenmiller, Hawk-moths of Vicinity of New York. 313 Pupa.—Robust, blackish brown, shagreened, terminal spine short, blunt and flattened horizontally; tongue-case concealed. Length. 1.80-2 inches=45-50 mm. Food-plants.—Willow and Poplar. Rather scarce in this vicinity, but more common in the North- ern and Western States. It is found from the Atlantic to the Pacific. It varies in depth of the ground color from light to dark, and in the distinctness of the transverse markings. In the Western States the prevailing form of this species is quite pale, and has been named occidentalis. In this neighborhood it is found during the latter part of July and August, and is probably double brooded. Smerinthus Zafrez//e. Head small, sunken, tufted between the antennz ; palpi short, not closely applied to the head ; tongue rudimentary ; eyes small ; antennz rather strongly pectinate in the male, simple in the female ; thorax stout ; abdomen more or less tufted at the sides. Fore wings more or less excavated outwardly or scol- loped ; anal angle produced ; inner margin excavated ; hind wings rounded, costa straight or excavated ; anal angle more or less produced. Smerinthus geminatus Say. PLATE VII, Fie. 4: Fore wings ashen gray, or brown with a rosy tinge ; across the basal third is an angulated deep brown line, the angle being a little below the middle and pointing outwardly ; across the wings are also a number of transverse wavy lines, usually more or less distinct but sometimes quite confluent with the ground color ; from the angle of the basal line a velvety brown dash runs out- wardly to the transverse roseate or gray line, sometimes filling the lower half of this interspace. The terminal space is light or dark brown, forming a dis- tinct apical lunule. Hind wings rosy red, with buff colored outer borders, which are rarely pink, and a large black spot near the anal angle containing one or two small blue spots. Head and thorax ashen gray or ashen brown, the latter with a deep brown patch occupying the entire space between the patagia. Underside of fore wings rosy red at base, outer half brown with wavy trans- verse lines; hind wings brown and grayish, powdery, with transverse curved lines. Expanse, 1.80-2.75 inches=45-65 inches. Larva.—Body green, paler dorsally, with numerous granulations ; along each side are seven oblique bands of a pale yellow color except the last, which is bright yellow ; on the anterior segments is also a stripe on the subdorsum ; 314 Bulletin American Museum of Natural History. |Vol. VII, head triangular, green, granulated, with an oblique yellow stripe on each side, meeting at the apex. Caudal horn bluish. granulated; thoracic feet green ; spiracles red. Length, 2-2.60 inches=50-56 mm. Pupa.—Very similar to that of .S. excecatus, but smaller. Food-plants.—Cherry, wild and cultivated, Plum, Apple, Elm, Oak, Hazel, Hornbeam (Carpfinus), Ironwood (Ostrya), Birch, Willow, Poplar, Ash, ete. Rather common in this vicinity. It is double brooded, the first brood appearing in June and July, and the second in August. The moth is subject to considerable variation from light to dark shades on the fore wings. It also varies in the form of the ocellus of the secondaries. It is found from Canada to Virginia, and westward to Iowa. Smerinthus excecatus (4. & S.). Prare VIL, Fie. 4: Fore wings fawn color, with a pinkish tinge and darker shades and mark- ings; the basal third is fawn color with one or two more or less distinct wavy brown lines ; beyond this is an oblique dark brown shade running from the costa at the basal fourth to the hind margin near the angle, where it is broken by two or three small black spots. Across the outer fourth are three sinuous lines of the same color as the base of the wing, and in which are two narrow darker lines; outer part of wings dark with a narrow, wavy, light band or shade. Hind wings rose red with a large black spot containing a blue centre. Underside of fore wings rose color, outer portion partly ochreous, with pinkish broken lines. Hind wings ochraceous with pink transverse lines. Head and thorax fawn color, the latter rich velvety brown along the middle ; abdo- men ochreous above, pinkish fawn color at sides. Expanse, 2-3.80 inches= 50-95 mm. Larva.—Body granulated, apple green, paler above, with oblique, yellow lateral bands and a yellow subdorsal stripe broken by the bands. lead tri- angular, green with a yellow stripe on each side uniting at the apex, granula- ted ; caudal horn green; thoracic feet reddish brown, bases yellow. Some- times the body is more or less marked and spotted with red. Length, 2.20 inches=55 mm. ; Pupa.—Dark brown ; head-case rounded, corrugated ; wing-cases smooth ; thorax and segments punctured ; terminal spine corrugated, and _ sharply pointed ; tongue-case concealed. Length, 1.20-1.80 inches= 30-40 mm. 1895.| Beutenmiiller, Hawk-moths of Vicinity of New Vork. 315 Food-plants.—Wistaria, Cherry, wild and cultivated, Plum, Apple, Pear, Raspberry, Rose, Sfir@a, Elm, Oak, Hazel, Hornbeam (Carfinus), Ironwood (Ostrya), Birch, Willow, Poplar, Ash, etc. Common in this neighborhood, and is double brooded, appearing in June and July and again in August. The moth is very vari- able in color; sometimes the males are very much darker with an olivaceous shade, tinged with purplish. It is found throughout the eastern United States and Canada. It may be easily recog- nized by having the outer edge of the fore wings regularly scol- loped, and by the rose-colored hind wings with the eye-like spot. Smerinthus myops (4. & 5.). PLATE VII, Fic. 5. Fore wings rich brown with a lilac wavy line across the basal third and sev- eral across the outer fourth; on the costa before the apex is a small yellow patch, and another before the anal angle ; sometimes these two patches almost run across the wing in shape of a band between the lilac lines. Head, thorax and abdomen light or dark brown, thorax yellowish along the middle. Hind wings yellow bordered with brown along the costa and outer border; in the yellow area is a large black spot with a blue centre. Underside of wings yel- low, more or less marked with brown, and with the lilac outer lines of the fore wings partly repeated ; across the hind wings are two or three wavy, lilac lines. Expanse, 2—2.50 inches=50-62 mm. Larva.—Green, not granulated, with seven yellow oblique lateral bands, and one on each side of the head ; along each side of the back a row of red spots and another row near the spiracles ; caudal horn green. Length, 2 inches= 50 mm. Pupa.—Same shape as that of S. excecatus, but smaller and less stout. Length, 1.20 inches=30 mm. Food-plants.—Wild and cultivated Cherry. Rare in this neighborhood ; found during June and July. It is probably double brooded. The moth may be easily known by the deep brown upper wings and yellow hind wings with the black eye-like spot. The larva very much resembles that of S. excecatus, but is smooth instead of granulated. It is found from Canada to Florida, and west to the Mississippi. 316 Bulletin American Museum of Natural History. (Vol. VII, Smerinthus astylus (Yury). PLATE VII, Fic. 6. Fore wings ochraceous brown with lighter and darker shades ; across the wings, from near the base to the inner margin near the angle, is a.dark oblique line ; along the inner margin, beyond the base, runs a black shade terminating on a yellowish brown spot before the hind angle; across the terminal space is a lilac line ; on the costa before the apex is a subtriangular orange ochreous patch from which runs a darker shade, losing itself in the ground color about the middle of the wing; before the costal patch are two short lilac lines. Hind wings paler than the fore wings, containing a round black spot with a bluish centre. Head and thorax with a lilac tinge, the latter ochreous brown along the middle ; abdomen lilaceous with a yellowish ochreous shade along each side of the back, and a darker dorsal line more or less distinct. Underside of wings ochreous, outer portion darker ; on the fore wings the lilac lines from above are repeated, as are also the orange ochraceous patches, the one near the hind angle much the larger; hind wings with two transverse lilac lines across the middle followed by a bright orange ochraceous band, terminal space dark. Expanse, 2-2.50 inches=50-62 mm. Larva.—Yellowish green ; yellowish along the dorsal region ; head broadly marked with pink on each side, this color uniting at the apex ; on the first to the end of the second segment is a pink subdorsal stripe ; on each side of the body are seven oblique, broad yellow bands, which are broadly marked anteri- orly with pink on the dorsal region ; spiracles in a pink patch ; thoracic feet pink ; abdominal and anal legs pinkish outside ; caudal horn pinkish at base, tip yellow and minutely forked ; head and body covered with small white granulations. Length, 2 inches=50 mm. Food-plants.—Various species of Huckleberry and Rosacez. This rare species may be distinguished from its congeners by its plain color and markings. ‘The life history is imperfectly known. It is found from Canada to Pennsylvania, and probably also southward and westward. Cressonia G. & R. Head small, sunken, with a ridge in front; palpi in the male long and ex- ceeding the vertex, not closely applied to the head, separated and divaricate at the tip; in the female the palpi are much shorter ; tongue rudimentary ; an- tennz of male strongly bipectinate, simple in the female; thorax short and stout, slightly crested along the middle ; abdomen of male tapering, with tufts 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. 317 along each side in form of dentations, hardly visible in the female. Fore wings as long as the body in the male, longer in the female, broad, outer margin den- tate, inner margin sinuate. Hind wings rounded, dentate outwardly. Cressonia juglandis (4.& S$). Prare Vil, Fre.'s. Fore wings pale fawn color, sometimes with a pinkish tint, or sometimes with light or dark brown shades between the transverse markings and outer portion of the wings; across the basal fourth is a narrow brown line, and another across the basal third ; across the outer fourth are also two parallel transverse lines, curved at the costa and running to the hind margin, the inner line ending near the middle of the wing; before the inner line is sometimes a transverse shade in form of a line. Hind wings with two or three lines across the middle. Head, thorax and abdomen pale fawn color, thorax light or dark brown along the middle. Undersides light or dark ochraceous with the outer transverse line from above on the fore wings repeated ; hind wings with lines same as above. Expanse, 2-3.20 inches=50-80 mm. Larva.—Green or brown ; head triangular, apex quite pointed and _ bifid, much more so than in the larvze of Smerinthus, with white granulations ; body elongated, tapering gradually from the seventh segment to the extremity, granu- lated with white ; along each side are seven light green or whitish oblique bands, composed of granules; caudal horn brownish, covered with black spinules. Sometimes the body is spotted with pink. Length, 2.50 inches= 63 mm. Pupa.—Dark brown, almost black, rough, covered with short points ; head- case with four projections and also one on each eye-case; antennze-cases with a row of short pointed spines ; tongue-case buried ; last segments very much flat- tened beneath and compressed laterally ; anal segment with a flat, truncate projection ; last few segments encircled with rows of short spines. Length, I,20-1.50 inches= 30-43 mm. Food-plants.—Hickory, Walnut and Ironwood (Ostryza). Not rare, and double brooded in this vicinity. The first brood appears in June and the second in August. The species is sub- ject to considerable variation; some specimens are uniformly pale fawn color or ochraceous, with the transverse lines distinct, while other examples are more or less covered with dark brown so as to almost obscure the ground color and transverse lines. It is found from Canada to Florida, westward to the Mississippi and Texas. The larva may be known by its triangular head with the apex pointed and bifid. 318 Bulletin American Museum of Natural History. {Vol. VII, Synopsis of Species of Smerinthine. Amorpha. Large species : Outer margin of fore wings regularly scolloped ; abdo- men obtuse at apex. Fore wings pale gray, with a very broad olive gray median band ; hind «wines; Shaded with claret redssea ctrl ietn rae A. modesta, Smerinthus. Head small, sunken, tufted between the antennz ; fore wings scolloped or more or less excavated. Hind wings with an eye-like spot. Fore wings scolloped, fawn color with darker shades ; hind wings [Oils erase ee chneat ee eiake ts a cb Baa ok wcicns S. excecaus. Fore wings excavate ; gray with dark brown markings ; hind wings punk: in ‘the middle: iy. .c e-em nn olan a ee ee ... S. geminatus. Fore wings less excavate than in gemdnatus, dark chocolate brown ; hindiwineshyellow centrally ayers) eee S. myops. Fore wings with outer margin almost entirely straight ; color almost Uholbiolde (Xcel KONA Dacca meeuende Socogeoose.s2 ... S. astylus, Cressonia. Wings broad, outer margin dentate ; palpi of male long, ascending, and divaricate at tip ; very short in the female ; hind wings without eye-like spots. Pale fawn color, sometimes shaded more or less with dark brown, with narrow transverse lines.............- aretha asthe C. juglandis. . 1895.| Beutenmiiller, Hawk-moths of Vicinity of New York. She Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. EXPLANATION OF PLATES. PLATE II. 1.—Hemaris thysbe (Fabr.). Hate A ‘* var. uniformis (G. & k.). 3—- “ gracilis (G. & R.). 4—- “ diffinis (Bdv.). 5.—Amphion nessus (Cramer). 6.—Sphecodina abbotii (Swains.). 7.—Deidamia inscripta (Harris). 8.—Everyx cherilus (Cramer). g.— ‘* myron (Cramer). 10.— ‘‘ versicolor (Harris). PiaTeE III. 1.—Aéllopos fadus (Cramer). 2.—Deilephila lineata (Fabr.). 3.— s galii (Rott.). 4.—Theretra tersa (Linn.). 5.—Pholus pandorus (Hibn.). 6.— fe achemon (Dru.). 7.— 3 vitis (Linn.). PLATE IV. 1.—Dilophonota ello (Linn.). Male. = Ba + - Female. 3.—Phlegethontius quinguemaculatus (H aw.). 4.— S¢ carolina (Linn.). 5.— 7 cingulatus (Fabr.). 6.— rusticus (Fabr.). PLATE V. 1.— Sphinx drupiferarum A. &S. 2.— ‘“ kalmie A. &S5. 3.— ‘“ lucitiosa Cramer. 4.— ‘* gordius Cramer. 5.— ‘“ chersis Hiibn. 6.—Argeus labrusce (Linn.). Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Bigs Fig. Fig. Fig. PLaTE VI. 1.— Sphinx canadensis Bdv. 2— ‘ eremitus (Hiibn.). 3.— “ plebetus Fabr. 4.—Dolba hyleus (Dru.). 5.—Chlenogramma jasminearum (Bdv.). 6.—Ceratomia amyntor (Hiibn.). 7.— a undulosa (Walker). eA, WALL 1.—Lapara coniferarum (A. &S.). ce 2.— bombycotdes, var. harristi (Clem ). 3.—Smerinthus geminatus Say. 4.— os excecatus (A. & S.). 5.— a myops (A. & S.). 6.— ss astylus (Dru.). 7-—Amorpha modesta (Harris). 8.—Cressonia juglandis G. & RX. Buiretin A. M. N. H. Vor. VII, Prate II. f A MU es E607 Ei. Dre 1. Hemaris thysbe (Faér.). 6. Sphecodina abbotii (Swazz.). 2. s ne var uniformis (G. & 2#.). 7. Deidamia inscripta (Harris). 3 a gracilis (CG. & R.). 8. Everyx choerilus (Cramer). 4. “« diffinis (Bdv.). 9 “© myron (Cramer). 5. Amphion nessus (Cvavz.). 10. «* versicolor (Hayvris). Butietin A. M.N. H. LJourer. per I 2 26 4 . Aéllopos fadus (Cramer). . Deilephila lineata (Faér.). er galii var. INTERMEDIA (Kéy.). . Theretra tersa (Zzzz.). on Vot. VII, Prate III. . Pholus pandorus (/7#0.). “© achemon (D7z). be vitis (Zzz7.). Buttetin A. M.N.H. ¥ 7 eh | e aut} Ny ew" a eon 4 Lou TEL. DEL. 1. Dilophonota ello (Zzxv.). 2. ce ‘© (female). 3. Phlegethontius quinquemaculatus (Haw.). Aun + Vor. VII, Prare IV. . Phlegethontius carolina (L7xz.). ce cingulatus (/aér.). se rusticus (/aér.). ‘Ae, Buttetin A. M. N. H. Vot. VII, PLATE V. Cae L -Jovre. Det bi x. Sphinx drupiferarum (4. & S’. 4. Sphinx gordius (Cramer). 2. cs kalmize (A. & S.). F ‘© chersis (H76z.). f ‘© lucitiosa (Craver). 6. Argeus labruscee (Z7772.). > ButieTin A. M. N. H. 1. Sphinx canadensis (Bdz.). 2. ‘© eremitus (H76.). 3- *« plebeius (Faér.). 4. Dolba hylzeus (D7x). Vou. VII, Pirate VI. 5. Chlaenogramma jasminearum (4az.) 6. Ceratomia amyntor (//7é.). 7. ss undulosa (Wad/.). Butietin A. M. N. H. Vo. VII, Pratre VIL. er AU ial nt ay a ; i tad a ) i ‘rms LJourEL., vet 1. Lapara coniferarum (4. & S.). . Smerinthus myops (A. & S.). 2. e bombycoides yar. harrisii (C7Zev.). “e astylus (Dr ay Amorpha modesta (/avrr7s). 3. Smerinthus geminatus (Say). Cressonia juglandis (G. & #.). 4. ss exceecatus (A. & S.). on am Article IX.— FURTHER NOTES ON TRINIDAD BIRDS, WITH A DESCRIPTION OF A NEW SPECIES OF SVYNALLAXIS. By Frank M. CHapMAN. A second visit to Trinidad during March and April, 1894, while made largely for the purpose of collecting mammals, resulted in the acquisition of notes on birds which supplement those pub- lished in the preceding volume of this Bulletin.’ On this occa- sion | was accompanied by Mr. William Brewster, and after a brief visit to my former headquarters near Princestown, we be- came the guests of Mr. Albert B. Carr, on his cacao estate at Caparo, in the west-central part of the island, seven miles east of Chaguanas. The country here is not unlike that about the rest- house where previous collections were made, the primeval forest being broken only by cacao estates. These, however, are younger and smaller, the region having been settled within comparatively recent years. Probably for this reason certain birds, which are common in the clearings and cacao groves about the rest-house, are as yet comparatively rare or wanting on Mr. Carr’s estate ; for example: Vireo chivi agilis, Ramphocelus yacapa magnirostrts, Elainea pagana, Pitangus sulphuratus, and Tyrannus melancholicus Ssairapa. The month of April was passed in the mountains which form the northern coast of the island. On their northern or seaward side the bases of these mountains are indented by but few bays ; on their southern side, however, they are penetrated by numerous valleys. Our home was near the head of one of the most beau- tiful of these—the Caura Valley—about seven miles from its opening on the plains. Here we were the guests of Mr. J. E. Lickfold. 1*QOn the Birds of the Island of Trinidad,’ Bull. Am. Mus. Nat. Hist., VI, 1804, pp. 1-86. [November, 1895.) [321] 21 322 Bulletin American Museum of Natural History. |Vol. VU, Mr. Lickfold’s house is at an elevation of 500 feet, while the crests of the surrounding hills reach an average altitude of about 2000 feet. The locality has long been devoted to cacao growing, and the primeval forest has largely disappeared. Still there are many large tracts of first-growth timber within a few hours’ ride. While I visited them on several occasions, my experience was too limited to render valuable a comparison of the avifauna of the mountains with that of the lowlands; and I leave to future ob- servers the task of explaining the rarity of such common lowland birds as Glaucis hirsutus, Pygmornis longuemareus, Phaéthornis guyl, Galbula ruficauda, Rhamphastos vitellinus, and Pionus men- struus ; while the following equally common lowland species were not once observed : Ostinops decumanus, Cassicus persicus, Pipra auricapilla, Momotus swainsoni, Trogon (three species), Amazona, and Urochroma. On the other hand, Luphonia trinttatis and Calliste desmaresti were observed only in the mountains. In attempting to express my appreciation of the hospitality extended me, I am again impressed by the failure of words to convey a sense of either my indebtedness or gratitude. Mr. Carr and Mr. Lickfold not only placed their homes at our disposal, but assisted us in every possible manner. I am also under many obligations to Mr. F. W. Urich. NOTES ON SPECIES NOT OBSERVED IN 1893. Merula phzopyga (Cuv/.). WHItTE-THROATED THRUSH.— One female was taken at Caparo and another at Caura. They agree in color with a Venezuelan specimen. Euphonia trinitatis (S¢r7ck/.). Cravar.—Not uncommon in the mountains, but not observed in the lowlands. Calliste desmaresti Gray. WorrHtrss.—Observed only on the crests of the ridges in the Caura district, where it was not uncommon. 1895.| Chapman, Further Notes on Trinidad Birds. 323 Piranga hemalea S.& G. Rurous Tanacer.—A male in the plumage of the female, but with testes measuring about .18 in their longer diameter, was taken at Caura April 21. Legatus albicollis (Vze//.). BLACK-BANDED PErcHARY.— A male of this species, heard calling from a tree-top, was taken at Caparo. Chasmorhynchus variegatus (Gm.). BELi-pirp ; Cam- PANERO.—This bird was not uncommon in the forests at Caparo, and in the more heavily wooded districts of the Caura Valley I have heard three birds calling at one time. The notes of this species will be found described at length in an article by Mr. Brewster and myself in ‘The Auk’ for July, 1895. Synallaxis carri, sp. nov. Synallaxis cinerascens LEOTAUD, mec TEMM. Char. Sp.—Similar to Synallaxis terrestris Jard., but upper parts, wings, and tail darker, throat blacker, rest of underparts darker and more olivaceous. Description of Type (Coll. Am. Mus., No. 60,614, male, Caparo, Trinidad, March 27, 1894; I'rank M. Chapman). —Upper parts mummy-brown ;' exposed portion of the wing-quills and wing-coverts deep chestnut-rufous, tail decidedly darker ; central third of the feathers of the upper throat white, lateral third black ; feathers of the lower throat centrally buffy ; rest of the underparts bistre with a slight cinnamon tinge, the breast faintly streaked with cinnamon. Wing, 2.08 ; tail, 2.52 ; exposed culmen, .53 inch. The differences between this bird and a specimen of 5S. ¢erres- tris from Tobago are found in its darker coloration throughout, and especially in the restriction of the white of the throat. In this character it resembles the Colombian 4S. Zemosticta, from which species it may, however, be distinguished at a glance by its more olivaceous and less rufous color. The only specimen secured was killed on the ground in the forests at Caparo. It gives me pleasure to dedicate this species to Mr. Albert B. Carr, of Trinidad, not only as a token of my gratitude for his 1 Cf Ridgway’s Nomenclature of Colors. 324 Bulletin American Museum of Natural History. |Vol. VXI, assistance, but also in recognition of his knowledge of the Trini- dad fauna. Chetura cinereicauda (Cass.).—A common species at Ca- paro, where four species of this genus were more or less abun- dant—the present, C. c/neretventris lawrence, C. spinicauda, and C. polioura. Frequently all four would be circling above us at the same time. C. céwereicauda has not been before recorded from Trinidad, and this capture extends its known range from Southern Brazil. I have no other specimens of C. c7nerezcauda for comparison, but my eight specimens differ from six Yucatan examples of C. gaumer? as stated by Mr. Hartert (Cat. Birds Brit. Mus., XVI, p. 482). Lurocalis semitorquatus (Gv.).—'!wo birds of this species were procured at Caparo. They were observed more or less regu- larly feeding at dusk near the border of the forest, flying swiftly back and forth over a short circuit and within ten feet of the ground. They thus resembled both a Nighthawk and Whip-poor- will in their feeding habits. A single low, insignificant note, uttered in flight, was the only one heard. Celeus elegans (J/7i//.). YELLOW-HEADED WoODPECKER.— One of two birds seen at Caparo was secured. Falco rufigularis Daud. Rep-THROATED FaLcon.—A speci- men was taken by Mr. Brewster. Cancroma cochlearia Zin. Boar-Bitt.—One immature specimen, in rufous plumage, was killed by Mr. Carr. ADDITIONAL NOTES ON BIRDS OBSERVED IN 1893. Thamnophilus major albicrissus (A/dew.). Thamnophilus albicrissus RipGW. Proc. U. S. N. M. XIV, 1891, p. 481. Thamnophilus major CHAPM. Bull. A. M. N. H. VI, 1894, p. 40. In reviewing my paper on Trinidad Birds,’ Mr. Ridgway speaks of the “Omission of Formicarius [lege Thamnophilus| 1 Auk, XI, 1894, p. 172. 1895.| Chapman, Further Notes on Trinidad Birds. 325 trinitatis and F. [lege T.| albicrissus described by me in the Pro- ceedings of the U. S. National Museum, Vol. XIV, No. 871, p. 481.” These birds were not omitted, but having overlooked Mr. Ridgway’s separation of them, I included them both under the names of the Continental forms. At my request Mr. Ridgway has kindly loaned me the two specimens upon which his descrip- tions were based. Comparison of the type of Zhamnophilus albi- crissus with seven males from Trinidad and twenty males of true T. major from Brazil, apparently proves the Trinidad bird to be a race of the latter distinguished by its larger bill, whiter under- parts, narrower white edgings on the outer vane of the primaries, and narrower white bars on the rectrices. ‘The character of ”” “remiges entirely without white edgings,” given by Mr. Ridgway, appears to be a variable one, dependent probably upon age. Im- mature specimens with brown wing-coverts, like the type, have no white on the primaries, but fully adult examples have well- developed margins to these feathers. A male from El] Pilar, Venezuela, and also one from British Guiana, agree with Trinidad specimens, and it is probable that all birds from north of the Amazon should stand as 7hamnophilus mayor albicrissus (Ridgw.). Thamnophilus cirrhatus (Gm). Thamnophilus trinitatis RiDGW. Proc. U. S. N. M. XIV, 1891, p. 481. As stated above, Mr. Ridgway has aiso loaned me his type of Thamnophilus trinitatts. The characters assigned to this race prove evidently, in my opinion, to be due largely if not entirely to individual variation. Two of three Trinidad specimens have the back of the same color as Mr. Ridgway’s type, while the third agrees in coloration with a Demararan specimen. ‘The color of the underparts is also variable. ‘Trinidad specimens may average darker below, but a specimen from Demarara is fully as dark, if not darker, than one from Trinidad. Phaéthornis guyi (Zess.). Brin-pLanc. —Notes on the song-habits of this species, by Mr. Brewster and myself, may be found in ‘ The Auk’ for July, 1895, p. 207. 326 Bulletin American Museum of Natural History. (Vol. VII] Nyctibius jamaicensis (Gm.). Poor-mE-one.—In the paper just cited (p. 208), our experience with this species is given in detail, and also an admirable colored plate. As surmised, /Vyc- tibius proved to be the author of the ‘ Poor-me-one’ call generally attributed to the Little Ant-eater (Cyclothurus didactylus), | Article X.— DESCRIPTIONS OF NEW AMERICAN MAMMALS. By J. A. ALLEN. During the last few months the Museum has acquired several quite important collections of mammals, which will later form the basis of special papers. As several months will elapse before their publication, it seems advisable to publish in advance descrip- tions of the several forms contained in these collections which appear to be new. Lepus aquaticus attwateri, subsp. nov. ATTWATER’S SWAMP HARE. Type, No. 4744, ¢ad., Medina River, 18 miles south of San Antonio, Texas, May 8, 1894; H. P. Attwater.—Above pale buffy gray, heavily lined with black, paler on the sides, which are whitish gray with a faint tinge of buff ; median dorsal area more strongly tinged with yellowish, increasing slightly in intensity on the rump ; nuchal patch, the fulvous ocular region, the pectoral band, and the outer surface of fore and hind limbs many shades paler than in L. aguaticus ; ventral surface and inside of fore and hind limbs white, the fur ashy plumbeous basally. Measurements (from the fresh specimen by the collector).—‘‘ Nose to end of tail, 520; tail [to end of hairs], 83; hind foot, 105. Weight, 514 lbs. Con- tained 3 large embryos.” Ear from notch (measured from skin), 65. Skull.—Total length (from posterior edge of occipital crest to front edge of nasals), 87; basal length (posterior border of occipital condyles to front of intermaxillaries), 79 ; zygomatic breadth, 4o ; mastoid breadth, 32 ; least inter- orbital breadth, 32 ; length of nasals, 35 ; greatest posterior breadth of nasals, 18 ; length of lower jaw, 63 ; height at coronoid process, 37. In a former paper on Texas mammals (this Bulletin, VI, 1894, p. 171) reference was made to two specimens of an aquatic hare collected at San Antonio by Mr. Attwater, which on comparison with specimens from Louisiana and the coast of Texas (Mata- gorda Bay region) proved to be much lighter colored than the latter. Mr. Attwater has since sent four others, making a series of six, taken as follows: one in March, three in April, one in May, and one in June, Oncomparison with a strictly comparable [327] 328 Bulletin American Museum of Natural History. [Vol. VII, series from the vicinity of Lake Catharine, Louisiana, the contrast in color is very striking, the San Antonio specimens being many shades paler throughout, lacking almost entirely the rich rusty fulvous tint of the Gulf Coast specimens. ‘This is shown quite as strongly in a young specimen, apparently not more than three weeks old, as in the adults. Fortunately there is a specimen of nearly the same age in the Louisiana series, so that both young and adults of the two forms are available for comparison. In short, the difference between Z. aguaticus and ZL. a. attwatert is quite as strong as between the Atlantic coast forms of the Z. sylvaticus group and their representatives in the arid interior. I take pleasure in naming this strongly marked subspecies in honor of Mr. H. P. Attwater, in recognition of his intelligent and persistent efforts to extend our knowledge of the mammal fauna of Texas. His experience with this inland form of Swamp Hare is detailed in the following note. “Swamp Rabbits are becoming very scarce, being much less numerous than they were ten years ago. Those I have met with were found in the drift piles and old fallen tree-tops in the most tangled parts of the San Antonio and Medina river bottoms. When frightened from their hiding places and chased by dogs they take refuge in hollow trees and in holes in the river bluffs. The dogs seem to have more difficulty in trailing them than they do the Cottontails:and Jack Rabbits, the Swamp Rabbits often eluding the hounds by taking to water. I have seen them on several occasions swimming across the river while the dogs were hunting for them on the other side. I have not heard of their occurrence north of San Antonio, and Mr. Lacey has not met with them on the Guadalupe River in Kerr County.” Reithrodontomys australis, sp. nov. TRAzG HARVEST MOUSE. Very similar in coloration and proportions to adults of 2etthrodontomys Jongicauda in winter pelage from California, but larger. Adult.—(Type.) Above warm yellowish brown, sparingly lined with black, darker medially and lighter and more yellowish on the sides, but without a 1895. | Allen, Descriptions of New American Mammals. 329 distinct fulvous lateral line ; beneath ashy plumbeous with a faint wash of buff, giving the effect of soiled ashy plumbeous, the fur being plumbeous at base. Feet grayish; ears blackish, weli haired; tail sharply bicolor, dusky brown above, whitish below, quite hairy, but the annuli not wholly concealed. Total length (measurements from skin), 158; tail vertebrae, 80; hind foot, 18 ; ear from crown, 10. Skull.—Total length, 23; basal length, 20; greatest cranial breadth, 11 ; ’ greatest zygomatic breadth, 10; least interorbital breadth, 3.7; length of nasals, 8. Type, No. 422+, ad., Volcan de Irazu, Costa Rica, June, 1892 ; George K. Cherrie. sar: This species is based on a single specimen (sex not indicated), received from Mr. A. Alfaro, labeled as above. In coloration and in general external features it bears a surprising similarity to R. longicauda. In this connection it is of interest to recall Mr. Tomes’s record (P. Z. S., 1861, p. 284) of * Retthrodon longicauda’ from Duejias, Guatemala. Among other interesting mammals received from Mr. Alfaro, and by him kindly presented to the Museum, may be mentioned a good series of Geomys cherried Allen’ (this Bulletin, V, 1893, p. 337), described originally from a single specimen, which show that the white crown spot is a constant and normal character. He has also sent a single specimen of ZAchimys semispinosus Tomes (P. Z. S., 1860, p. 265), described from Ecuador, but since re- corded from Nicaragua and Costa Rica (Pacuare) by Mr. F. W. True (Proc. U. S. Nat. Mus., 1888, p. 467). The specimen is labeled ““Suerre,’ Costa Rica, alt. 1500 ft., July, 1895. A. Alfaro.” Oryzomys cherriei, sp. nov. CHERRIE’S CoTron Rat. Pelage rather coarse ; size medium ; tail rather short, considerably less than half the total length. Adult.—Above yellowish brown, varied with blackish tipped hairs, darkest along the middle of the back, lighter and grayer on the sides ; below whitish gray, the fur dusky at base and tipped with whitish. The color of the lower surface passes gradually into the grayish brown of the flanks. Feet and ears 1 Macrogeomys cherrtet Merriam, N. Am, Fauna, No. 8, 1895, p. 104, pl. xv, fig. 1. 330 Bulletin American Museum of Natural History. |Vol. VU, gray; tail nearly naked, indistinctly bicolor—dusky brown above, lighter, grayish brown below. Half-grown young are wholly plumbeous below, and darker and less washed with yellowish brown above than adults. Measurements (average of 16 adults, 1044, 6 22).—Total length, 214 ; tail vertebrz, 92 (collector’s measurements from fresh specimens). Hind foot, 23; ear from crown, 12 (measurements from skins). Skull,—The skull differs from that of O. palustris in no very important feature except in being much smaller. Total length (occipital plane to front border of nasals), 30; basal length (occipital condyles to front edge of inter- maxillaries), 28; greatest zygomatic breadth, 16; greatest breadth of brain- case, 13 ; least interorbital breadth, 6. Type, No. 4!22;', 6 ad., Boruca, Costa Rica, Dec. 10, 1891; George K. Cherrie. Based on a series of 21 specimens (16 adult, 5 juv.), collected at Boruca, Costa Rica, Nov. 1g-Dec. 10, 1891, by Mr. George K. Cherrie, for whom the species 1s named. Oryzomys cherriei needs comparison with no other species known to me. In general appearance it most resembles O. palus- tris, but it is fully one-third smaller than any of the known forms of this species, from all of which it also differs decidedly in colora- tion. It has no close relation to any other described Central American species of the genus. Peromyscus attwateri, sp. nov. ATTWATER’S CLIFF MOUSE. Above tawny brown, darker and much mixed with blackish along the median dorsal area, more golden on the sides, the lower edge of the dorsal area forming a strongly defined golden lateral line. Below pure white, the base of the fur plumbeous. Fore feet white to slightly above the wrists; hind feet white nearly to the tarsal joint, soles naked nearly to the heels. Ears very large, nearly naked, dusky, faintly edged with whitish. ‘ail sharply bicolor, dusky above, grayish below, moderately well haired (the annulations showing through more or less towards the base), and generally well tufted at the end. Measurements.—Average of to adult specimens, measured in the flesh: Total length, 196 (187-216) mm. ; tail vertebrae, 100 (96-110); hind foot, 21 (20-23) ; ear from notch (measured from the skins), 16 (15-17). The type, a breeding female, is rather above the average of the series, measuring as fol- lows: Total length, 216; tail vertebra, 110 ; hind foot, 23 ; ear, 17. 1895.] Allen, Descriptions of New American Mammals. 331 Skull (of type), total length, 28 ; basilar length, 26; greatest cranial breadth, 14; least interorbital breadth, 5 ; length of nasals, 9.5. Type, No. 1°A4%', 2 ad., Turtle Creek, Kerr Co., Texas, March 12, 1895; H. P. Attwater. This species is based on a series of 14 specimens collected on Turtle Creek, Kerr Co., Texas, May 24, 1894, and March 9-13, 1895, and on 3 from San Geronimo Creek, Medina Co., Texas, April 3, 1895. Several are in the nearly uniform dark gray pelage of the young, others are more advanced but still immature, while about one- half are ‘ young’ adults, only a few being ‘old’ adults. One only (the type) has a very small spot of bright fulvous on the breast. Peromyscus attwateri finds its nearest affines in Peromyscus row- leyi and P. evemicus, but seems to be clearly different from either. This species is named for the collector, Mr. H. P. Attwater, who contributes the following interesting field notes. “T call these mice ‘Cliff Mice,’ to distinguish them from the other form (7. mearnsii), because they are found in large numbers in the cracks.and cavities of the rocky cliffs that border the rivers and smaller streams in the counties directly north and west of San Antonio. ‘Though most numerous along the sides of canons, they are also found in hollow trees, logs, fences, and cultivated fields, and about ranch buildings in the valleys, as well as in the cedar ‘ brakes’ on the divides and high ground. “The southern limit of the range of this species, in this part of Texas, is about ten or twelve miles north and west of San Antonio. I have not found it at San Antonio or south of it, and do not think it will be found east of Bexar or Comal Counties. The short tailed form (?. mearnsi?) doubtless extends up the valleys into the range of the Cliff Mice, but the latter seem to restrict themselves to the rocky country. “They feed on the different nuts and seeds which grow in end- less variety all over this region, though their favorite food seems to be acorns and cedar berries. I believe they also prey exten- sively on birds’ eggs.” Neotoma cinnamomea, sp. nov. FuLtvous Woop Rat. Similar to VV. rupicola Allen, but larger, coloration much deeper, and the ears darker. 332 Bulletin American Museum of Natural History. |Vol. VII, Adult,—Above, in summer, buffy-ochraceous, with often a tinge of vinaceous ; middle of dorsal region finely lined with black ; sides clear strong ochraceous buff ; feet and ventral surface pure white to the base of the hairs ; tail bushy, dusky gray above, pure white below; ears brownish, thinly haired. Young.—The young in first pelage are ashy above, with a tinge of fulvous, conspicuously varied with black, especially over the middle of the dorsal region ; below white, with a tinge of ashy along the sides of the abdomen, owing to the slight plumbeous cast of the underfur. ‘Tail terete, ashy white above, a little clearer white below. From this stage they pass into the autumn coat, in which the upper parts are cream buff with an ashy shade, strongly lined with black ; below white, with the basal portion of the fur on the sides of the abdomen ashy or pale plumbeous. Tail colored nearly as in the adult, but much less bushy. Measurements.—Votal length (average of 6 adult males), 364 (356-368) ; tail vertebrae, 158 (151-163) ; hind foot, 41 (40-43) ; ear from notch (measured from dry skins), 27 (25-28). Four adult breeding females, average slightly smaller, as follows: Total length, 343 (337-351) ; tail vertebrae, 148 (144-150) ; hind foot, 39 (37-41) ; ear from notch (from dry skins), 27 (26-28). Type, No. 41992, 6 ad., Kinney Ranch, Bitter Creek, Wyoming, July 9, 1895 ; Walter W. Granger. This species is based primarily on a series of 31 specimens, collected by Mr. Granger at Kinney Ranch, Bitter Creek, Wyo- ming, July 6-Aug. 6; to which are also referred 2 specimens taken on the Uncompahgre Indian Reservation, Utah, April 2 and 9, and 3 taken on the Little Snake River, near the Colorado- Wyoming boundary line, Aug. 26. The adults of both sexes are well represented, as are the immature stages, from quarter-grown young to full-grown young of the year. This species belongs to the same group of bushy-tailed Wood Rats as JV. orolestes Merriam and JV. rupico/a Allen, being inter- mediate between them in size, but quite different from either in coloration. It is much smaller than ZV. ovolestes, which it appears to most resemble in color. It differs from WV. rupicola in its considerably larger size, and in its much deeper and more vina- ceous buff shade of coloration, and much darker ears at all ages. Microsciurus, subgen. nov. Skull short, broad, the dorsal outline very convex, postorbital processes placed slightly behind the widest part of the malar, which is remarkably expanded. Premolars 7. 1895.| Allen, Descriptions of New American Mammals. 333 In regard to external characters, the tail is narrow, and the ears are very small; the pelage is short and close. Type, Scvurus (MWicroscturus) alfart, sp. nov. This group of Squirrels, which will probably be found to include all of the Guerlinguets (as Sc7wrus pusillus Desm., and 5S. chrysuros Puch.), is exceptional for the peculiar form of the malar and the relatively great breadth and convexity of the skull. Sciurus (Microsciurus) alfari, sp. nov. ALFARO’S SQUIRREL. Total length (measurements from dry skins), 290 ; head and body, 145 ; tail vertebrze, 105 ; tail to end of hairs, 145 ; hind foot, 35 ; ear from crown, 9. Above, including upper surface of both fore and hind feet, dark olivaceous brown, minutely punctated with yellowish rusty, the hairs being blackish, slightly tipped with yellowish rusty, giving a dusky olivaceous general effect, becoming, however, more reddish brown on the head ; below and inside of limbs fulvous gray, varying in different specimens from buff to strong fulvous, and even rufous. Upper surface. of tail uniform in color with the back ; lower surface similar, hence much darker than ventral surface of the body, with which it is in strong contrast. The hairs of the lower surface of the tail are individually dark red- dish brown or deep chestnut, with three narrow bands of black, the outer much broader than the others. Whole front of head washed with dark rufous, strongest on the sides of the head. Ears small, rounded, showing but little above the surrounding pelage. In two old females the space enclosing each nipple is gray. Skull.—Total length (front border of nasals to occipital crest), 36; basal length (front border of intermaxillaries to occipital condyles), 32; greatest zygomatic breadth, 22 ; least interorbital breadth, 13; nasals, ro. Type, G2e¢, 9 ad., Jimenez, Costa Rica, Jan. 24, 1894; George K. Cherrie. Based on three females, two of which are adult, and the other about two- thirds grown, collected by Messrs. Anastasio Alfaro and George K. Cherrie. This species should be compared with Scéwrus pusillus Desm. and S. chrysuros Puch., from either of which, judging from de- scriptions, it differs quite markedly in color, and from the former also in size. Tamias pricei, sp. nov. PRICE’s CHIPMUNK. Intermediate in size and coloration between 7. merriami and 7. hindsiz, but very different and about equally distinct from either. 334 Bulletin American Museum of Natural History. {Vol. VU, Breeding Pelage (April specimens).—General color above, dull grayish brown, or gray varied with hazel and brown. Flanks tawny; sides of shoulders and thighs strongly grayish ; lower surface whitish, the abdominal area washed more or less strongly with dull yellowish brown ; color of the flanks often encroaching considerably upon the sides of the ventral area. Dorsal stripes nine—five mixed hazei and black, and four clear ashy gray. The median dark stripe extends from the nape to the base of the tail ; the anterior third is mainly brown, mixed more or less in different specimens with black ; the posterior half or two-thirds mainly black, edged and more or less varied with hazel. The first lateral dark stripe on either side is similar to the median one, but is shorter, extending generally only from the shoulder to a little beyond the hip, and contains less black. The outer dusky stripe is still shorter and only slightly varied with black. Inner pair of light stripes gray ; outer pair broader and clearer gray. Post auricular patches small, dull grayish white ; light facial stripes clear gray ; the dark ones dull hazel brown, lighter than in 7. Azvdsiz, but much darker than in 7. merriami. ‘Tail above pale, the color beneath the surface being clay color, which shows conspicuously through the surface, the individual hairs being black at the extreme base, and then ringed broadly and about equally with clay color and black and tipped with whitish gray; tail below centrally deep reddish chestnut, with a narrow border of black fringed with gray—about as in 7. hindsiz, Ears of medium size (much smaller than in 7. merriamz), externally blackish on the anterior portion and gray on the posterior third or half. Measurements.—Average and extremes of 23 646: Total length, 252! (234-278) ; tail vertebrae, 119 (109-130) ; hind foot, 35 (32-37). Averages and extremes of 17 99: Total length, 256 (241-271); tail vertebra, 122.5 (113-130) ; hind foot, 35 (32-37). Type, No. 4283, ¢ ad., Portola, San Mateo Co., California, April 12, 1895 ; 9552) J. Diefenbach. Named for Mr. W. W. Price. This very distinct form of Tamas is based on a series of 45 specimens taken at Portola, in the Santa Cruz Mountains, Cali- fornia, during the last week of March and the first two weeks of April, by Messrs. R. L. Wilbur and J. Diefenbach for Mr. W. W. Price, to whom [I am indebted for the opportunity of examining a large collection of mammals from different localities in the Santa Cruz Mountains. Zam/as priced is almost exactly intermediate in all essential features between 7. Azndsit of the coast region of California north of San Francisco and 7. merriami of the moun- tains of southern California (San Diego, and San Bernardino 1 From nose to end of caudal vertebrze; about 30 mm. should be added for the extension of the hairs beyond the vertebra. 1895.] Allen, Descriptions of New American Mammals. 335 Counties, etc.). The gap between 7. prvcei and these forms is so evenly balanced that it is difficult to say to which of them 7. priced is most closely allied. The line of separation from either, so far as present material goes, is so sharp that it seems best for the present to treat the new form as specifically distinct from either, although it seems not improbable that specimens from interme- diate points between the present known ranges of the three forms may show their complete intergradation. As at present known, T. pricet is much more distinct from either 7. merriami or 7. hindsit than 7. obscurus is from 7. merriamt, or than 7. townsendit is from 7. Aindsi. Tamias wortmani, sp. nov. WoORTMAN’S CHIPMUNK. Female, Breeding Pelage.—Above dull yellowish gray, with a slight vinaceous tinge. A narrow yellowish white band on either side from the ear to the hip, with no dusky band (or only a very slight trace of one) on the zzmerside of the white band ; a short broad black band on the outside of the white band; sides of body below the black band yellowish white ; sides of neck and shoulders scarcely more yellowish than the sides of the body ; ventral surface whitish or grayish white, the dusky basal portion of the pelage more or less visible through the surface ; tail above grizzled dusky and pale yellowish, the hairs being black, tipped and sub-basally ringed with pale fulvous ; tail below, buff, with a broad subapical zone of black, and a narrow line of black near the base of the lateral hairs, visible only on parting the hairs ; feet buffy gray. In several specimens the lower, as well as the usual upper, black band on the sides of the body is wholly wanting. The male is probably similar, but doubtless a little brighter colored, especially on the sides of the shoulders. (The males when taken had already assumed the post-breeding pelage.) Male, Post-breeding Pelage.—Middle of the dorsal region, from the nape to the tail, yellowish gray, varied with black-tipped hairs, and with a faint wash of vinaceous, bounded on either side by a broad line of yellowish white, extend- ing from the shoulders to the hips; top of head more strongly vinaceous or tufescent ; sides of neck and shoulders deep ochraceous, cutting off the lateral white line at the shoulders. Below the white lateral line is a short broad band of deep black ; sides of body behind the shoulders straw yellow ; below buffy white, with a tinge of dusky, due to the dusky basal portion of the pelage show- ing through the surface. Tail above dusky, edged and varied with fulvous ; below pale fulvous, with a submarginal narrow black band, and a narrow dusky line at the extreme base of the lateral hairs, only seen on carefully parting the hairs. 336 Bulletin American Museum of Natural History. (Vol. V1, The female at this stage is probably similar but paler, especially on the sides of the shoulders. (None of the adult females when taken had acquired the post-breeding dress.) Young in First Pelage.—Pelage soft and thin; above dull yellowish gray- brown, with a well-defined narrow white lateral line, and below this a short, broad dusky band ; sides of body and below grayish white ; tail above grizzled fulvous and dusky, below fulvous centrally, submarginally dusky, and edged with pale fulvous. Young in Molt.—A large series of young of the year show that the young molt directly from the first pelage into a dress similar to the post-breeding pelage of the adults. This series also shows that the females are much less richly colored than the males, particularly over the sides of the neck and shoulders. Measurements.—Seven adult females give the following averages and ex- tremes: Total length, 280 (271-292) ; tail vertebrae, g5 (87-100) ; hind foot, 42 (41-44). Four adult males give the following : Total length, 272 (260-282) ; tail vertebrae, 96 (92-101) ; hind foot, 43 (42-44). . Type, No. ‘p03, 4 ad. (still partly in molt), Kinney Ranch, Bitter Creek, 935 Wyoming, July 13, 1895; Walter W. Granger. Named for Dr. J. L. Wortman, the Director of the American Museum Palze- ontological Expeditions to the western Bad Lands, to whose interest in Mr. Granger's work is largely due his eminent success during the field seasons of 1894 and 1895. This species is based on a series of 55 specimens collected by Mr. W. W. Granger, in the vicinity of Bitter Creek, Wyoming, July 5-Aug. 2, 1895. The series consists largely of young of the year, but includes about a dozen adults, representing both sexes. Both adult and young are in molt, but the greater part have nearly acquired the post breeding dress. This species is perhaps most nearly related to Zamzas lateralis, but combines in a singular manner the characters of the two ‘couplets’ into which Dr. Merriam, a few years since (N. Am. Fauna, No. 4, Oct., 1890, p. 18), separated the 7° datera/is group. It differs from 7. Zateralis and 7. cinerascens in having the whole under surface of the tail (except the submarginal black band common to all) uniform fulvous as in 7. castanurus and 7. chry- sodetrus, but differs from the latter, and also from 7. /aferalis and T. cinerascens in the entire absence of the inner black lateral band. In the entire series of 55 specimens, made up of examples of all 1895.| Allen, Descriptions of New American Mammals. 337 ages and conditions of pelage, only three or four show even a trace of this inner black dorsal band—a feature sometimes lack- ing, it is true, in 7. Zateralis ; but in 7. wortmani its nearly uniform absence is combined with a fulvous lower tail surface. Spermophilus tridecemlineatus olivaceus, subsp. nov. BLACK HILLS SPERMOPHILE. Similar in size and markings to S. tridecemlineatus pallidus, but much darker, as regards the ground color of the upper parts, with the light stripes and spots pale yellowish olivaceous. Breeding Pelage (July females).—Above ground color dusky brown or black- ish, with no trace of ferrugineous or chestnut ; flanks, stripes, and spots pale creamy buff with a slight olivaceous effect ; below rather strong cream buff. Post-breeding Pelage (July males).—Pelage longer and softer, but coloration not appreciably different. Young of the year are similar in coloration to the adults. Measurements.—Averages and extremes of 7 breeding adults(2 44 and 5 29): Total length, 252 (245-265) ; tail vertebrz, 89 (76-94) ; hind foot, 34.5 (33-37). Type, No. 23%, 2 ad., Custer, Black Hills, South Dakota, July 25, 1894 ; W. W. Granger. This strongly-marked subspecies is based on 7 adults and 12 young of the year, the latter one-half to two-thirds grown. It differs from fa//idus in its much darker ground color and the olivaceous creamy white tint of the light stripes and spots. Spermophilus tridecemlineatus parvus, subsp. nov. SMALL STRIPED SPERMOPHILE. Much smaller than either S. tridecemlineatus or S. ¢t. pallidus, and very different in coloration from either. Breeding Pelage (April specimens).—Coloration, especially the ground color of the dorsal surface, much paler than in either S. tridecemlineatus or S. t. pal- fidus. (Ground color above deep russet, slightly varied with blackish; the stripes and spots grayish white with a very faint tinge of cream color ; feet and ventral surface white, the dusky basal portion of the hairs showing faintly through the surface. Measurements.— Average and extremes of 5 specimens: Total length, 204 (200-207) ; tail vertebrae, 80 (75-86) ; hind foot, 30.6 (30-31). Type No. 12335, 4 ad., Uncompahgre Indian Reservation, northeastern Utah, May 2, 1895; W. W. Granger. [ December, 1895.| Zz SS 338 Bulletin American Museum of Natural History. |Vol. VU, This subspecies is based primarily on two specimens from the Uncompahgre Indian Reservation, in northeastern Utah, taken respectively April 4 and May 2, to which are also referred 8 specimens from the vicinity of Bitter Creek, in southwestern Wyoming, taken July 21-Aug. 5, all collected by Mr. W. W. Granger. The series is very uniform in coloration, the exception being one or two of the July specimens in which the light stripes and spots, and also the flanks, are slightly more tinged with a faint wash of creamy white. The type of Mitchell’s Scéwrus tridecemlineatus came from “ the sources of the Mississippi River,” and hence from Central Min- nesota. Professor Baird, writing in 1857 (Mam. N. Am., p. 317), observes that specimens “ from Wisconsin are seen to differ quite materially from those [from] further west, in a considerably larger the spots and lines,” he adds, being “not so large in proportion as in the lighter prairie specimens.” In 1874 I separated (Proc. Boston Soc. Nat. Hist., XVI, 1874, p. 291) the pale western form here referred to as Spermophilus tr1- decemlineatus pallidus, without, however, giving any diagnosis. 90 66 size and darker color, This was supplied three years later in my monographic revision of the American Sciuride (Mon. N. Am. Roden., 1877, p. 873). As was customary at the time, no type was designated, but it was stated that ‘‘ Among the smallest and palest examples are the specimens from Fort Union and the Yellowstone and Platte Rivers, an especially pale and small phase characterizing the Mauvaises Terres of the Upper Missouri.”’ In now separating additional forms of this group, | would restrict pal/idus to the arid region of the Plains, from the Upper Missouri southward to eastern Colorado, western Kansas, etc., and designate as its type region the plains of the Lower Yellow- stone River. The four forms of Spermophilus tridecemlineatus here recog- nized may be diagnosed as follows, the characters being based in each case on breeding specimens : Ground color of dorsal surface blackish chestnut—dark chestnut mixed with a profusion of black-tipped hairs, the black generally prevailing ; light stripes and spots pale yellowish white, the light stripes less than half the width of the intervening dark spaces; lower parts buffy white, the hairs dusky oF cy 10h el eae enn coarctation Aaia0.0 « ... tridecemlineatus. | | se te gi ty 1895.| Allen, Descriptions of New American Mammals. 339 Ground color above clear chestnut, scantily varied with black-tipped hairs, the prevailing tone being rather light chestnut ; light stripes and spots creamy white, the light stripes being nearly as wide as the intervening dark spaces ; lower parts pale yellowish white to the base of the hairs...... .. pallidus. Ground color dusky yellowish gray, the dark tint being ade up of an intimate mixture of yellowish gray and black-tipped hairs, generally wholly without chestnut or ferrugineous; light stripes and spots pale yellowish white with a tinge of olivaceous buff ; below pale creamy white, the hairs eA Geimaimtne: DASE ese, eee. tit Scie tics ek eae ews cece olivaceus. Ground color russet, sparingly varied with black-tipped hairs ; stripes and spots grayish white with a faint creamy tinge ; lower parts clear whitish MET AV icetal Siete) ela MCi ays emi wa aiie al alnlwicelel alee rela dete Ce cia mee vine r ew nee parvus. Measurements. Total length. | Tail vertebra. | Hind foot. S. tridecemlineatus'.. .. | 293 (283-314) | 99 (90-112) 40 (38-42) ME PONIAUS, =. vou os | 227 (203-260) | 73 (61- 89) 32.5 (31-34) S. t. olivaceus®..........| 252 (245-277) 89 (76— 94) 34.5 (33-37) BEL POT UUS smi ach. sioyss © | 204 (200-207) 80 (75— 86) 30.6 (80-31) Blarina (Soriciscus) nigrescens, sp. nov. Blarina micrura ALLEN, Bull. Am. Mus. Nat. Hist., V, 1893, p. 338, not Sorex micrurus TOMES (=8larina micrurva ALSTON), from Duejfias, Guatemala. Pelage coarse, rather long, and not lustrous. Above dusky plumbeous, in some lights black; lower surface not appreciably different. Feet and tail blackish, nearly naked, the annulations of the latter being distinctly visible. Measurements.—Head and body, 65 ; tail vertebrae, 22 ; hind foot, 12. Skull, total length, 20 ; mastoid breadth, 9.5 ; length of nasals, 7; length of upper tooth row, g ; distance between outer edges of last molars, 6.3. Type (and only specimen), No. 2224, adult, San Isidro (San José), Costa Rica, Sept. 5, 1891 ; George K. Cherrie. This species is of about the size and proportions of Sorex [7. ¢., Blarina| micrurus Tomes (P. Z. S., 1861, p. 279), described from Bectas, Guatemala, but it is obviously very different in coloration, 1 Fort Sablling. Nas Pear are: 38 er adgie Geedine specimens, measured in the flesh by Dr. E. A. Mearns, Uz a i ae Allen, Mon. N re Roden., p. 877—16 specimens, various localities, probably not all adult. * Custer, Black Hills, S. D.; 244, 5 92 —all adult breeding specimens. + Bitter Creek, Wyo. ; 5 specimens—all practically adult. 340 Bulletin American Museum of Natural History. |Vol. V11.] B. micrura being described as having the upper parts “ darkish grey-brown, with a slight grisly appearance,” and the “whole under surface” as “ lightish grey-brown, tinged on the chin and along the middle of the abdomen with yellowish rufous,” with the feet and tail “of a lightish grey colour.” Blarina (Soriciscus) orophila, sp. nov. Pelage glossy, very short, soft and velvety. Above dark brown (shading slightly on seal brown), becoming lighter on the sides, and passing gradually into smoke gray on the ventral surface, where the hairs are conspicuously tipped with whitish. Feet grayish brown ; tail dusky above, distinctly lighter below, well clothed, and with a minute pencil at the tip. Ears rudimentary and not easily detected. Measurements.—Head and body, 55; tail vertebra, 21; hind foot, 11 ; head, 20. Skull (too imperfect for complete measurements).—Length of nasals, 5 ; length of upper tooth row, 8 ; distance between outer borders of last molars, Ee Type, No. $$49, adult, Volcan de Irazu, Feb., 1894 ; George K. Cherrie. Based on two specimens, preserved in alcohol, one of which is adult, the other immature, collected as above, and kindly pre- sented by Mr. Cherrie to the Museum. ‘The description is based on the specimens in a dry state, after removal from alcohol, to which they have since been returned for safer preservation. In color this species somewhat resembles Alarina cinerea but it is very much darker, and has a much longer tail. It is nearly one- third smaller (in actual bulk) than either 4. mecrura or B. nigrescens, and very different in color from either, particularly the latter, to which it has, for this group of animals, comparatively no resemblance. Blarina orophila differs very strikingly from &. nigrescens in the entire structure of the first upper molariform tooth, the first outer cusp of which rises to the same height as the others, instead of being rudimentary and uncolored, as in 4. zigrescens. There are also differences in every detail of structure between the two teeth, and also in the structure of the third upper molariform tooth, in the two species. Article XI.— NOTES ON SOME SPECIMENS OF MINERALS FROM WASHINGTON’ HEIGHTS, NEW YORK CITY. By E. O. Hovey. Recent excavation at 171st Street and Fort Washington Avenue, New York City, has brought to light mineral specimens of more than local interest on account of their rarity or size or both, and the purpose of this article is to put on record some facts concern- ing their dimensions and occurrence. The minerals occurred in three pockets close together, aggregating eighty feet (about 24 m.) in length in a vein of coarse pegmatite in mica schist. ‘The strike of the schist is about N. 30 E. (magnetic), and its dip 85° east- ward, and the vein is essentially parallel thereto, with a maximum width of about 3 feet (1 m.). The body of the vein is granular gray quartz, feldspar (orthoclase), and flaky muscovite, in which were imbedded the specimens noted in the following lines. To Mr. William Niven is due the credit for discovering and exploit- ing the deposit. Xenotime.—During the excavation of the roadbed of the new speedway along the Harlem River Mr. Niven’ found a great many small xenotimes imbedded in oligoclase, and a few large ones, one of them being the largest ever found on the island up to that time. At the locality at present under discussion the xenotimes were fewer in number but averaged larger in size, and one exceeded the largest found on the speedway. The last mentioned crystal is 8 -+- mm. square, and its approximate height is6 mm. _ It is a simple symmetrical octahedron composed of the unit pyramid. The color is clear yellowish brown, and the crystal is imbedded in granular gray quartz. A second crystal is 5.5 mm. by 6 mm. in horizontal dimensions, and has a semi-altitude of 3mm. It is composed of the unit pyramid predominating with the unit prism well developed, and a second pyramid ¢ (311) indicated. It is imbedded in feldspar and mica. The third xenotime to be noted is a very perfectly preserved one imbedded 1 Vid. On a new locality for Xenotime, Monazite, etc.,on Manhattan Island. Am. Jour. Sci. III, 1, 75, 1895. [341] 342 Bulletin American Museum of Natural History. |Vol. Vi1.| so as to show only one set of pyramidal faces. It is 7-++mm. long and 5 mm. wide with an apparent semi-altitude of about 3 mm., and is surrounded by the three minerals of the vein. The planes of all these crystals are pitted as if by etching. Monazite.—This mineral, of good quality, was found in numerous small crystals and parallel growths. ‘The largest crystal is translucent, clove-brown in color, and very perfect in its development. It is imbedded in quartz and feldspar. The portion exposed measures 13.5 mm. long and 6.5 mm. wide; the whole length may be 18 mm. ‘The crystal is strongly columnar in habit and is not flattened on the orthodiagonal axis, as is so commonly the case in monazite. ‘The planes present are a (110), n (120) and + (111) predominating, 7 (110) narrow and w (ror) narrow and interrupted. A detached group of smaller crystals showing in addition to the planes just mentioned, the two clino- domes, ¢ (o11) and w (021), is ro mm. in total height and 6 mm. in diameter. All the planes are more or less pitted. Tourmaline.—Black tourmaline was abundant in the vein, mostly in small, brilliant crystals, but there were some large ones, of which one is worthy of note. It is 243 mm. long and 96 mm. in greatest diameter, the least diameter being 80 mm. It is a very simple crystal showing one termination consisting of the rhom- bohedra, ~ (torr) and ¢ (o112). The body of the crystal is quite round, but on one side it has a small parallel growth, without terminal planes, extending about two-thirds of its length. At 164.5 mm. from the terminated end a seam of granular gray quartz from 5 to 9 mm. thick divides the crystal into two portions, but otherwise it is very compact. ‘The matrix is granular gray quartz. Miscellaneous.—Other minerals occurring here, in association with those already mentioned, are zircon in small long prismatic crystals, dumortierite, torbernite (? ), autunite (?), apatite, musco- vite, and garnet (almandite). The apatite is green in color, abundant, and is sometimes seen in small perfect crystals pene- trating the tourmaline. As further indications of the large scale upon which the minerals crystallized at this locality, it may be mentioned that there were found several very large aggregates of garnets in parallel position, and crystals of muscovite fifteen centimeters and more in longest diameter. ‘The largest garnet is about half of a single crystal which would measure 23 centi- meters in axial diameter if it were complete. It is a trapezo- hedron with the dodecahedral planes well developed. The dumortierite occurs not only in the feldspar, but also as long filiform inclusions in the muscovite, singly and radiating from centers, Article XII. — PERISSODACTYLS OF THE LOWER MIOCENE WHITE RIVER BEDS. By HENRY FAIRFIELD OsBorN and J. L. WortTMAN. With Plates VIII-XI and twelve Figures in the Text. INTRODUCTION. The progress of our knowledge of the White River fossil fauna has been extremely rapid since 1892, owing chiefly to the discovery of the ‘ Protoceras Beds,’ the location of the “ Metamynodon level,’ and the very exact stratigraphic and expert collecting methods employed by the American Museum and Princeton exploring parties. The most welcome result of the field work is that we are now securing complete skeletons of animals which have been hitherto represented only by isolated skulls and limbs. We can now replace the useful but largely conjectural ‘restorations’ of the last decade by figures taken directly from the skeletons. The two types illustrated in this Bulletin are the massive 77tanotherium, and the smaller but no less interesting J/etamynodon, drawn from complete skeletons which have recently been mounted by Mr. Hermann for the new hall of Vertebrate Palzontology. The second result, less striking perhaps, but of equal importance, is that we are obtaining very much more perfect and abundant examples of the rarer forms of White River Mammals. The present paper is confined to the publication of new or little known characters of the Perissodactyla, and includes the following points of chief interest : 1. The entire skeleton of 77tanotherium robustum is described. The vertebral formula is shown to differ from that of all other Perissodactyla, and to agree with that of the Artiodactyla. It is probable that certain wide differences in the development of the 1343] 344 Bulletin American Museum of Natural History. \Vol. VU, horns, which have been assigned a generic value, are merely sexual characters. 2. The White River Horses exhibit a very marked evolution in size as we pass from the lower to the upper White River levels. There is apparently a direct specific succession connecting Mesohippus bairdii Leidy, through AZ. tntermedius (nobis) of the ‘Protoceras Beds,’ with Anchitherium prestans Cope of the John Day Beds. A distinct, very much larger, and apparently new type of Horse is the JZ. coper of the Protoceras Beds. We have thus in the horse line reached the point long ago pre- dicted by Lamarck in promulgating the evolution theory, namely, that the lines drawn in the Linnean system of nomenclature would be finally obliterated by discovery. In fact we are now beginning to retain the binomial system upon grounds of con- venience and of scientific courtesy, rather than upon lines of definition. 3. The true Lophiodontidz of Europe are found to be repre- sented in this country by the AMeptodon-Helaletes-Colodon \ine previously referred by Marsh and ourselves to the Helaletide. The alleged Ayrachus douvillet of Filhol is actually identical with Colodon, showing that a contemporary transformation of the Lophiodons occurred in Europe and in this country. The peculiar foot formerly referred by us to AZesohippus longipes now appears to belong to a member of this phylum. 4. The differences between the Tapir, Lophiodon and Hyra- chyus molar types are clearly defined. 5. The skull of Hyrachyus agrarius from the Bridger Beds is described in this connection. 6. The mounted skeleton of MMZetamynodon planifrons is described. tactyls of White River. 345 er lS SOl 1895.| Osborn and Wortman, P GEOLOGICAL SUCCESSION. The stratigraphical position of the species described in this ing table : he follow in t hown in is § Bullet “WINnjsnqod UNLoYyjoUuR J, Hpate snddiyosa jy ‘wnpouosiy wnieyyei.0y ‘snjepidsnooid ‘> ‘sisuajoyep uopofoy ‘ttpareq snddiyosayy ‘atu winliayywasoy ‘suojrueyd uopoudweja ‘st[eJUapIo00 uopojo,) ‘ttp.req snddiyosayy ‘areyuap -1990 WINLeyyeiooy ‘azLW Winey eV ‘yaput ‘ds ‘uop -oudwejayy ‘ayejuapio90 ~wnLeYyIeia0Vy ‘tadoo snddiyoso jy ‘snipoutiajur snddiyosayy ‘wunyeydao -Qed sy ‘wnyAjorpiy = wntiayjeid90y ‘spoq ssdd -snddiyosa |. ‘WM LIoYyye1l0y ‘WNLIaYyOURIL], JO SuleWal polsulyy | ‘auYyM souog ‘Avo A143 ysippay “Aysna souog = *sAvyjo pue sauojspuvs :4sade7y uopoudwejo AOA] PA, “ePLXO snouLsnasay jo ojvos YUM souog ‘SsuLiwaq-aynpou : 194v"y UOpoaid ‘pasojoo ‘sAv[D pure soauojspues Ajsna souog ‘uinjyeays Avjo snoynpoN ‘O}IM Souog “SARI PolO[OO-}Y SI] *SHONUTPUOS JOU ‘sauoyjspues 951807) ‘sAvjo pasojoo-xurd ‘SuLvaq-ajnpou :19Xe7y vruayonejday ‘YoY jo jOVAVYD Bauer) el (1aay Ogr ‘ssaUAIIYI [¥I0,],) ‘saqaq WOINWHLONVEITL *jo0} OOL ! ) ‘J99} OZ 0} OF ‘saagq NOGOdNO *JoaJ OOI 0} SL “SAVTD) NUWUVE ‘yooy SL-oS ‘Saag SVIAOOLOU ‘J99] OOT “spog ay) Jo ssauxorya aya jo ajyeunse ajeuntxoiddy ‘NoDaNGO ‘saaq AVC NHO[ ‘VLOYVG HLNOS ‘sadag AIAIY ALIN AA 3460 Bulletin American Museum of Natural History. |Vol. VII, Family TITANOTHERIID Zé. Genus Titanotherium Zed). Titanotherium robustum J/ars/. PLATES VIII AND IX. The chief result of the Museum Expedition of 1892, under Dr. Wortman, assisted by Mr. O. A. Peterson, was the discovery of a large 7Trtanotherium skeleton (No. 518) in the upper Titanotherium beds of South Dakota near the head of Corral Cafion. The skull was first found, in a somewhat shattered condition, and then the neck, entire trunk and fore limbs, perfect even to the sesamoids, were excavated as far back as the last lumbar vertebra and the border of one ilium. At this point, to their great disappoint- ment, the party encountered a sudden change in the rock, and found that the sacrum, remainder of the pelvis and hind limbs had been carried away by an erosion which had _ probably occurred some time after the original deposition of the entire animal. A vigorous search in the summer of 1894 for hind limbs of the proper proportions resulted only in the finding of a left tibia (No. 1075) and fibula (No. 1071), and a left pes (No. 1073), left caleaneum and astragalus (No. 1076). Finally, by the kind codperation of the Princeton parties under Mr. Hatcher and Mr. J. W. Gidley, the Museum secured a perfect pelvis (No. 1065) and two femora (Nos. 1442, 1443) belonging to three different individuals. The size of these parts was deter- mined (1) by the fact that the pelvis corresponds very closely to that belonging to the main skeleton; (2) one of the femora had associated with it metacarpal bones, which also agree in size with those of the main skeleton. We thus have every reason to believe that the proportions between the fore and hind limbs are very nearly accurate. The entire animal was then put together and mounted with the greatest skill by Mr. Adam Hermann, head preparator of the Department of Vertebrate Paleontology. The only parts which he found it necessary to restore were the teeth of the left side and certain smaller gaps in the skull; the sacrum and a few of the 1895.] Osborn and Wortman, Perissodactyls of White River. 347 ro i‘ » / caudals ; the cuboid, navicular and cuneiforms of the left pes ; part of the right tibia (No. 493), caleaneum (1073), and the major part of the right pes. These missing parts were carefully modeled from the opposite side or from other individuals of different size. The only parts missing are the manubrium sterni and some of the posterior sternals. — Fig. 1. 7itanotherium robustum. Mounted skeleton (No. 518), ¢ , seen from three-fourths front view. Approximately one-thirtieth natural size. The completed skeleton is about 14 feet long, 8 feet high and 4 feet broad. The teeth are well worn, yet the epiphyses upon the summits of the dorsals indicate that the animal was not fully adult. An interesting feature of the skeleton is the exostosis and false joint in the center of the seventh rib, undoubtedly an after result of fracture. The skeleton differs from the Scott-Osborn restoration of Titanothertum proutit (Fig. 2) mainly because 7. proutit is a more primitive and less robust type. Marsh's restoration of 7. (Brontops) robustum,’ executed by Mr. Berger, is a remarkably 1 Am. Journ. Sci., Feb., 1889. p. 163. 348 Bulletin American Museum of Natural History. |Vol. VII, skillful drawing of the trunk and limbs, but errs in the too small proportions of the skull, as seen by a comparison with our Fig. I, a perspective drawing by Mr. Weber. ‘The Scott-Osborn and Marsh restorations are both at fault, however, in placing too many vertebre in the dorso-lumbar series. This animal actually possessed but ¢zwenxdy dorso-lumbars. Fig. 2. Vvtanotherium prouti?. As restored in 1887 by Scott and Osborn; now modified by reduction of the lumbars, One-thirty-second natural size. SEXUAL AND SPECIFIC CHARACTERS. This animal was found in the same level (Upper Titanothe- rium Beds), and agrees closely in size and appearance with the type skeleton of Zztanotherium (Brontops) robustum of Marsh.’ The cheek teeth characters (pm.=4, m= 3) are also the same. In the American Museum specimen the premaxillaries are imper- fect, and we cannot determine the number of incisors in either jaw. In Marsh’s type there are two upper incisors. In the Museum collection there is also a fine skull (No. 492) with a very 1 Am. Journ. Sci., Oct., 1887, p. 4. 1895.| Osborn and Wortman, Perissodactyls of White River. 349 long pair of horns. This agrees closely with Marsh’s type of T. ( Titanops) elatum.’ It is noteworthy that the alleged distinct species, Z. robustum and 7. elatum, and the American Museum specimens similar to them, both occur upon the same level, and were therefore contemporaneous. They agree in the length of the nasals, in many minor details of skull structure, and in the characters of the dentition. They differ mainly in the szze of the horns, a character which is very generally of sexual signifi- cance only. The conclusion appears very probable that the genus and species Zittanops elatus is founded upon a male individual of the genus and species Brontops robustus, the latter having been established upon a female individual. The species 7. robustum appears to differ somewhat from the previously established 7. dolichoceras Scott and Osborn,” in the flatter horn section and longer nasals, but it may subse- quently appear that these differences are not of specific value. Dimensions of Skeleton. Feet. Inches.| Metres. Length, tips of nasals to bend of tail................. he 29.236 4.15 “SIRI RG ONI¢ 5 2s Se coer a ay 0a ae oe He fe 2530 Me AtEL MAGFOSSPCLViS: |. . sc. eae s+ So dims ct oO ieds 3 Io 1.18 Hind limb, total length Fees inte SE ISE See Ce ee 5 6 1.67 POSTED, oS | RS me eno i a aaa -79 amass Po hie tiere Rab 0 Coie COE Sea ae eee 1 4% -42 Metatarsal ITI, length. . Ue ates So eae eee 8 205 Fore limb, total length, including ne 5 Sea eee 6, 9 2.05 Scapula go TE So sar oy SRO ORE» en ee 2 246. | 67 |S ner FG) Se See ee ee ee I 94 55 ECE TTS © tlie tine snes ee kei opal ges a a (Pe a .46 ina; including oleeranon= Wy. hy. He sw ss ke ss I 1144 .60 Mietacarpalv EME lencthy i. foe 2 a4 San ses ss. s 9 23 Skull, length, incisors to condyles ..........-.. .. a a ls .8o Molars, Premolars, Canine inclusive .............. Teas 45 Vertebral column, total length, excluding caudals, (including intervertebral spaces).............. gq 386°] 7 Cervicals, total, inferior centra.. Besta. te yaitys és DA tek Srp fe SALS ee eee SARE PS ne oS. os 5 5 1.65 3 Lumbars, “s Be oe ae ro .28 4 Sacrals, 2S Gainrteal eee 5 eee 7% | 19 20 Caudals, ‘ He eR e8 aS. a ae 3 946 | 1.15 4th Dorsal Vertebra, length, with spine ............ 2 314 70 5th Rib, length, outer measure ...........-.. Sei 2 Br | .g6 Sth > x COS hei cee ti Ae eee 3. 76 fa Dake 1 Op. cit., p. o. * Bull. Mus. Comp. Zodél., Vol. XIII, 1887, p. 160. 359 Bulletin American Museum of Natural History. |Vol. VU, The most characteristic features of the animal are the following : Skull.— 'The nasals are of medium length; the horns are short, forwardly projecting, and imperfectly ossified at the tips. The zygomatic portion of the squamosal shows a decided poste- rior bulge but no shelf-like projection. The supra-occipital border is deeply indented. Vertebre.—TVhe fine series of vertebree belonging to No. 518, complete to the last lumbar but lacking the sacrals and caudals, enables us to fully describe and illustrate the backbone. The plate (Pl. [X) is taken from an enlarged drawing made just after the vertebree were mounted. ‘The exceptional number of dorso- lumbar vertebrae suggests the note that a fracture was found through the center of the first lumbar, but there is no probability that one of the lumbars is missing. The formula is: Cervicals, 7 ; dorsalis, 175 himbatrs, 3 -sacralsie4: The number of dorso-lumbars therefore coincides with that in the Artiodactyla, namely, D.L. = 20, and is from three to four less than that typical of the Perissodactyla, namely, D.L. = 23-4. This corroborates a view already advanced by Osborn,’ that of all Perissodactyla the Titanotheres present the greatest number of affinities to the Artiodactyla; these affinities may now be sum- marized-as follows: the artiodactyl type of fore foot, the artio- dactyl type of superior molars, the vertebral formula character- istic of the Artiodactyla. It is premature to infer more from these facts than that if the Artiodactyla and Perissodactyla were derived from a common stem form, as expressed in the larger division Diplarthra of Cope, the Titanotheres have diverged less from this stem than other Perissodactyls, at least in the develop- ment of the above-mentioned characters. It is possible also that the shortening of the backbone may be secondary, so that the above generalization requires further verification by the discovery of the vertebral formula in the ancestral ‘Titanotheres. In details the vertebrae show many resemblances to those of Paleosyops paludosus, as described by Earle. ‘The atlas has a broad powerful transverse process with an inferior flange pierced by the vertebrarterial canal; the suboccipital nerve issued just above the anterior border of the process. ‘The axis has a peg-like '* Rise of the Mammalia in North America,’ p. 34. 1895.| Osborn and Wortman, Perissodactyls of White River. 351 odontoid and a powerful spine. The cervicals 3-6 are charac- terized by a progressive increase in the height of the neural spine, in the size of the transverse process and extension and depression of its inferior lamella; the post-zygapophyses are flat, similar in shape, and face downwards and outwards. The 7th cervical is Fig. 3. 7%tanotherium robustum. Mounted skeleton seen from behind. Approximately one-thirtieth natural size. imperforate with a greatly reduced transverse process. The dorsals are characterized by the sudden elevation, in d. 1-4, and gradual sinking of the spines as we pass backwards. Every dorsal from d. 1-f7 is characterized by a facet for both the head and tubercle of the corresponding rib. The zygapophysial facets he in a nearly horizontal plane from d. 1 tod. 11; they then gradually shift to an oblique plane from d. 12 to d. 14; and into a nearly vertical plane in d. 15-16. ‘The zygapophyses of the 17th dorsal and rst lumbar vertebrz are distinguished from all the others by being slightly concavo-convex. The post-zygapophysis of the 2d lumbar is plane and slightly oblique in position. The lumbar metapophy- 352 Bulletin American Museum of Natural History. |Vol. VIL, ses are flat and horizontal. The 3d lumbar articulates by an oblique facet and broad metapophysial process with the 1st sacral. The sacrum is unfortunately missing. Marsh states that there are four in this species.’ We find four in the perfectly preserved pelvis (No. 492) associated with the supposed male skull. The caudals are from a number of different individuals. The neural spines apparently extend back to the 8th vertebrae. The transverse processes die out upon the 6th. A well-developed chevron appears upon the 2d, and perhaps in a perfect series would be found upon the 3d. Fore Limb.—The fore limb is of an extremely robust character. The scapula shows a projection of the anterior border, a rounded and rugose superior border, and a long incurved posterior border. The most striking bone is the humerus with its huge plate-like great tuberosity, strong deltoid ridge, and powerful ectocondylar ridge. The shaft of the ulna is trihedral in section and stands well out from that of the radius. The radius has a flattened shaft and a well-marked inferior extensor groove. The structure of the manus is typically pavaxonic or artiodactyl, the median axis of the foot lying between the third and fourth digits. Other features of the skeleton are well illustrated in the drawings. Family EQUID. Subfamily ANCHITHERIINZ, Genus Mesohippus Jars. Representatives of this genus are exceedingly abundant in the White River formation, and as a result of the several expeditions made by the Museum party into these beds an unusually fine series of Horses of this epoch is contained in the collection. Several definitions of the genus have been given, the latest of which is by Scott,* in which he assigns the presence or absence of the enamel pit in the superior incisors to distinguish it from the John Day Horses, which he places under the generic title of Miohippus. He ascribes to Mesohippus complete absence of any enamel invagination in the upper incisors, but adds in a footnote, 1 Am. Journ. Sci., Feb., 1889, p. 1€4. 2 ‘Trans. Amer. Philos. Soc., 1893, p. 70. 1895.] Osborn and Wortman, Perissodactyls of White River. 353 “The upper incisors of this genus are not known, and future discovery may show that it is not generically different from JZ7o- hippus, but the generally less advanced character of the dentition renders it probable that the character of the incisors is as assumed above.” There are in our collections two specimens in which the supe- rior incisors are preserved in an almost perfect condition ; they both show a very decided pitting of the enamel in the two outer teeth. It will therefore be readily seen that the generic distinction between the White River and John Day species fails, and we really know of no characters of generic value by which they can be distinguished. In a like manner the distinctions between Mesohippus and Anchitherium disappear when one examines carefully a large series of White River and John Day Horses. Previous to the discovery of the Protoceras fauna in the upper part of the White River beds, but a single species, A/. bairdit, had been generally recognized’ in this formation, but with the acquisition of a large amount of material from the upper level it is now possible to demonstrate that there were two and probably three species living in that region when the successive sediments were laid down. -y\ Synopsis OF SPECIES OF MESOHIPPUS. M., batrdit. M. intermedius. M. copet. 1. Median pair of incisors | 1. Median pair of incisors | 1. Unknown. not cupped. slightly cupped. 2. Length of median| 2. Length of median| 2. Unknown. metapodial of fore- metapodial of fore- foot, .o80-.095. foot, .130-.132. 3. Length of median) 3. Length of median) 3. Length of median metapodial of hind- metapodial of hind- metapodial of hind- foot, .107—.124. foot, .I5I-.152. foot, .189. 4. Parastyle of Sup. Pm. | 4. Parastyle of Sup. Pm. | 4.” Parastyle of Sup. Pm. 2, small. 2, enlarged. 2, slightly enlarged. 5. Intermediate cusps of 5. Intermediate cusps 5.° Intermediate cusps of Sup. Ms. and Pms. | same asin M/.bairdit. Sup. Ms. and Pms. little separated from well separated from internal cusps. internal cusps. 6. Length of tibia, .240. | 6. Length of tibia, .317. 1 Several species have been proposed for remains from this horizon, but it seems probable from the Be enpsons that they pertain only to individual varieties of the most prevalent species 17. bairdii. Marsh has deccriee d M. celer,and Cope has described MW. cuneatum and M. exoletum from the Miocene of Colorado. 2 These characters are taken from the second specimen, No, 683. [ December, 1895.| 23 354 Bulletin American Museum of Natural History. |Vol. VII, Mesohippus intermedius, sp. nov. This species is based upon an almost complete skeleton (No. 1196) from the sandstones of the Protoceras layer of White River. There are, moreover, numerous other specimens including perfect feet, skulls, jaws and other parts of the skeleton from the same layer of both the White and Cheyenne River localities contained in the collection. These specimens all agree very closely in size, and average nearly one-third larger than JZ. dacrdiz from the lower or Oreodon layer. A comparison of the length of the median metapodials in different individuals is as follows : M. bairdti. M. intermedius. ( M. M. | .107 ee Length of median metapodial, hind foot......... 4.114 4.151 feeendy, ee | .124 .152 \ M. { M. Length of median metapodial, fore foot ........ + .080 + .130 | .095 | .132 It will be seen from this table that there is marked increase in the size and length of the metapodials of JZ. bairdiz, and it is interesting to note that the smallest examples of the species in our collection at least come from the lower layers, while the largest examples were found in the highest levels of the Oreodon stratum. Not only do our specimens of J. éatrdii show great varia- tion in size, but marked ¢vadividual variability in important struc- tural characters as well. Fully fifty per cent. of the specimens show coéssification of the three cuneiforms into a single bone ; others have the middle and internal cuneiforms united, while others again have all three bones free. The degree of reduction of the lateral metapodials is subject to much variation, as is also the extent of the development of the metapodial keels. The teeth vary greatly in the details of their structure, some showing much greater advancement than others. In MW. intermedius the variation is apparently not so great, especially as regards size. In some specimens the metapodials a eh aglretinrhy a ——————— 1895.] Osborn and Wortman, Perissodactyls of White River. 355 are thicker and stouter, the lateral ones being subcircular in sec- tion near the middle, while in other specimens the metapodials are decidedly more slender, the lateral ones being highly com- pressed laterally and very elliptical in cross section. In contrast Fig. 4. Right hind foot and left fore foot of Wesohippus intermedius, front and side views. P. pisiform, Zz. lunar, sc. scaphoid, #z. magnum, «#7. unciform, cw. cuneiform, cé. cuboid, 7. navicular, c3. external, c?. middle cuneiform, c/ facet for fibula. Slightly less than one-third natural size. with MW. dairdii the arrangement of the cuneiform bones seems to be very constant ; the middle and internal are always united, while the external is free. Another important distinction between J/. batrdii and M. intermedius is seen in the degree of the cupping of the incisors. In M. bairdii the two outer incisors are very distinctly cupped, 356 Bulletin American Museum of Natural History. [Vol. VII, but the median pair show no traces whatever of the enamel pit. In MZ. intermedius, on the other hand, the median pair are slightly but distinctly cupped. In this respect the incisors of JZ. inter- medius stand exactly half-way between those of JZ. dazrdiz and the John Day species, Anchitherium prestans, in which the median incisors are always distinctly and almost as strongly cupped as the two outer ones. In the superior premolar dentition there are also important differences which point strongly in the direction of the John Day species, especially Anchitheritum prestans. In M. bairdii the internal cingulum of the first superior premolar is but little developed, and does not form with the principal cusp a distinct basin; in WZ. zvtermedius the cingulum is more strongly developed and a distinct basin is formed. In the second superior premolar of JZ. dacrdiz the parastyle or cingular cusp at the antero-external angle of the crown is small and scarcely larger than those on the succeeding teeth. In JZ. intermeatus this cusp of the second premolar is considerably en- larged, giving to the crown an incipient triangular appearance. In Anchitherium prestans the enlargement of this cusp is carried still further, and in Protohippus and Eguus the crown of the tooth is of a triangular shape in front. The chief distinctions between MZ. intermedius and Anchithe- rium prestans are seen in the cupping of the median pair of incisors, the greater enlargement of the parastyle of the second superior premolar, the union of the posterior cross-crest with the outer wall in the superior molars and premolars, the greater reduction of the lateral metapodials, and the larger size of the latter species. Mesohippus copéi, sp. nov. This species is founded upon the complete half of a pelvis, femur, tibia, and part of a hind foot (No. 1197), together with a complete median metapodial, and one lateral metapodial of the hind foot of another individual (No. 1198), a collateral type. These remains indicate an animal much larger than JZ. znter- medius, and this is, so far as we know, the largest horse of the White River epoch, larger even than A. prestans of the John . ——— eee 1895.] Osborn and Wortman, Perissodactyls of White River. 357 Day. A comparison of the measurements of these bones with those of JZ. intermedius is as follows : M. copet. M. intermedius. M. M. Beret A ROpmtlld ata cp yan /4 lies eves ee leis ciecice ss ByiX5, .240 \ Nita hs egos Ye ¥S irae cor TINS ee Cee .O41 .035 ILigavedilal 82° 9 ge Ce ee a eee ee .048 .O41 Length of middle metapodial of hind foot..... .189 151 UME OO ee CLV enon ece.siolaP eps cininle erst, qed aie ti stause 334 PUENTE Ns bare c cisely vd evecare ve. a a ee 231 There are also in our collection two superior premolars (No. 683) of the right side, apparently the second and third of the series, that are much larger than any specimens of MV. interme- dius. We have therefore provis- ionally referred these teeth to this species. If this reference is correct, these teeth indicate a species quite different structur- ally from JZ. intermedius. Be- ‘sides their greater size, the inter- mediate cusps are much more distinct, being separated from ps p2 i : Fig. 5. Second and third right upper the internal cusps by a wide, premolars of Wesohippus copez. Crown view. , Ms. mesostyle, 4s. hypostyle, fs. protostyle. deep notch, whereas in JZ. Natural size. intermedius they form with the internal cusps a high crest and are very little separated. The measurements of these premolars are as follows : M. copet. M. intermedius. M. M. Length of second and third superior premolars... .037 .030 Widthtotsecondipremolat.s.- 2: 2. =) =. ae ZOLS O15 Width of thirdspremolar.--.. 2. 4. CMEC 202i .O17 This species differs from Axchitherium prestans in the less reduced character of the lateral metapodials, and in the lack of completion of the cross-crests of the superior premolars, as well as the distinctness of the intermediate cusps. ‘The two species are nearly equal in size. All of our material is from the Protoceras layer of the Cheyenne River locality, but a large foot, probably of this species, was found by Mr. J. B. Hatcher, of the Princeton expedition, in the Oreodon Beds. 358 Bulletin American Museum of Natural History. [Vol. VII, GEOLOGICAL SUCCESSION. gree | & = | <| John Day Beds. A. prestans, =} ~ 4 fo) S a a) o S fo) §) 4 ‘< Protoceras Beds. M. intermedius. M. copet. ASE ills ee g 150 feet. NS 4 . oo . = 8 % Oreodon Beds. M, bairdii. ? M. cope. S os 140 feet. awe = Titanotherium Beds. M. bairdii. Q) 180 feet—total. , | | The above table represents the nearly continuous sedimentation from the Titanotherium Beds into the John Day, having a total thickness of about eight hundred feet. _ There can be little doubt that the three types, Mesohippus bairaii, M. intermedius and A. prestans, form a distinct and closely connected phylogenetic series of animals slowly special- izing and constantly increasing in size. So far as we know ¢here ts not a single character missing tn the structural chain. Meso- hippus or Anchitherium copet, on the other hand, is somewhat larger than A. prestans, and forms a side branch, leading possibly into, one of the numerous parallel species which Cope and Scott have described from the John Day and Deep River Beds. Family LOPHIODONTID. (Sensu strictu.) A family of lophodont Perissodactyls intermediate between the Tapiridz and Hyracodontide. Superior molars, with paracone and metacone of same size but differing in shape. Metacone pushed inwards, more or less concave. Paracone lengthened. Metacone shortened. Fleptodon. | — Lophiodon. Flelatletes. Colodon, Incisors ,, pre- Incisors 3, pre- Incisors , pre- Incisors », pre- molars }-4. Third molars 3, without} molars 4. Third) molars }. Second, and fourth superior posterior crests. | and fourth supe- third and fourth premolars without) Manus and_ pes|rior premolars with superior premo- posterior crests. | unknown. posterior crests. | lars with posterior Digits 4-3 Median | crests. toes enlarged. | 1895.| Osborn and Wortman, Perissodactyls of White River. 359 It now proves that Leidy was very near the truth in referring to Cuvier’s genus Zophiodon certain Bridger (ZL. nanum) and White River (Z. occidentalis) jaws and teeth. The discovery of the superior molar series of Co/odon demonstrates beyond a doubt f \ \ . ‘ i ‘ r t . / se parastyle paracone metacone parastyle paracone metacone parastyle paracone metarone \ " / \ i i \ é , ‘ \ / i ; ° \ \ ' 1 é crisla ! t Z t i \ protoioph metaloph protoloph metaloph 7 \ i \ Xs I \ \ Za ; metalophid hypolophid —_hypoconulia metalophid hypolophid htt metalophid hypolophid Tapiroip. LopHIoDONT. HyRACODONT. (Systemodon.) (Lophiodon.) (Hyrachyus.) Fig. 6. Principat LopHioponr Morar TyPEs. that true Lophiodontide, not in the loose sense of the term of Cope, Lydekker and Flower,’ but in the strict phylogenetic or true relationship sense, were represented in North America by the animals hitherto grouped in the family Helaletide by Marsh and Osborn. This family identity has been anticipated by Osborn.? The true American Lophiodonts are now seen to be 1 “Mammals, Living and Extinct,’ 1891, p. 373. By these authors, Hyracotherium, Syste- modon, Hyrachyus, in fact all lophiodont Perissodactylsin which the premolars are simpler than the molars, are termed ‘Lophiodonts’ without regard to the wide gaps which separate them from the true Lophiodon. 2* Fossil Mammals of the Wahsatch and Wind River Beds,’ Bull. Am. Mus., 18g2, p. 92. Also ‘Rise of Mammalia in North America,’ 1893, p- 39- 360 Bulletin American Museum of Natural Fiistory. |Vol. VU, Heptodon of our Wahsatch, He/aletes of the Bridger and Uinta, and Colodon of the White River. It now appears that besides Lophiodon, both Helaletes and Colodon probably occur in Europe as the last representatives of the Lophiodon line. Cuvier’s type, Z. tapirotdes, is a lower jaw found at Issel,’ an horizon which contains Pachynolophus, and is approximately equivalent to our Bridger. The Bridger species of /e/aletes, namely : H. (Hyrachyus) nanus Leidy, H. boops Marsh, H. (Des- matotherium) guyotit 8.& O., H.(Dilophodon) minusculus 8. & O., are well known to differ from the Issel Lophiodons (Z. ¢apiroides Cuvier, Z. ¢ssedensis Fischer) in the possession of rudimentary transverse crests upon fwo superior premolars. In the higher White River horizon the species of Colodon, namely, C. occiden- talis Leidy, C. (2) longipes O. & W., C. dakotensis O. & W., C. pro- cuspidatus O. & W., differ still further from ZLophzodon in the pos- session of posterior crests upon ¢ree of the upper premolars. The true molar pattern in Heptodon, Lophiodon, Helaletes and Colodon is identical; the question arises, can we separate the oldest American type, the Wind River or basal Bridger Hleptodon, with its unmodified premolars, from Lophzodon? It now seems that we can do so. So far as we know, Cope’s Heptodon is nearly identical with Cuvier’s Zophiodon, the only distinction being one | of size, and the number of upper premolars. The likeness is in the identical pattern of the molar teeth and the absence of pos- terior crests upon the premolars. The skeleton of Hef/odon, as previously shown by the writers,” is highly specialized, resembling that of the Hyracodons in many respects, but tending still more to monodactylism. ‘The climax of this tendency is shown in a White River hind-limb, which we at first® referred to JZesohippus, but which now appears to belong to a form probably related to Colodon. The extremities of Lophio- don are not known, or have not been described. The nearest approach to the Heptodon type of skeleton in the French Eocene beds is that which has been referred to Paloplotherium minus by the French paleontologists. The ?. minus tarsus and hind limb are almost identical in size and in numerous minor characteristics 1 Ossem. Fossils, 2d edition, Vol. if: \P- 176, p 2 ‘Fossil Mammals of the Wabhsatch,’ Bull. ‘Mus., Vol. 1V, Sept., 1892, p. 131. 3 Bull. Am. Mus., 1894, p. 214. My ttt a (14 WN eG ch RI Mt Bi a 1895.| Osbornand Wortman, Perissodactyls of White River. 361 with the Heptodon limb. We do not know whether the associa- tion of the P. minus skeletal parts with the teeth of the Pa/oplo- thertum type is absolutely demonstrated ; if it is not, it seems quite probable that the so-called P. mznus feet belong not to the Palzotheres (from which they differ so widely), but to some small Lophiodont such as Hefptodon. Genus Heptodon Cope. For a full account of this Wind River type, see our paper upon the Wahsatch Fossil Mammals, and Prof. Cope’s description in the ‘ Tertiary Vertebrata.’ Genus Lophiodon Cuvier. Under this genus should be included only those forms with simple premolars which are ¢dentical in molar pattern with Cuvier’s type, such as Z. ¢apirotdes Cuvier, L. tsselensis Fisher, L. paristense Gervais, L. buchsovillanum Blainville. We may confidently exc/ude all those European forms which have the true Tapir, Rhinoceros, Hyracodon or Amynodon molar pattern, and which undoubtedly belong to animals ancestral to Cadurcothertum, to Protapirus, to Aceratherium, and possibly to the Hyracodonts. This will remove from Lophivodon a host of wrongly-referred species. The question, What is Lophiodon?’ seems now nearer solu- tion. It is intermediate in molar pattern and in skeletal charac- ters between the Tapirs and Hyracodons or Rhinoceroses, and shows a mingling of their characters, but represents a line of descent entirely distinct from both. Genus Helaletes Marsh. For the synonomy and characteristics of this type, see Scott and Osborn’s Memoir upon ‘Mammalia of the Uinta Formation,’ our paper upon the Wahsatch Mammals,’ and Wortman and Earle’s paper upon ‘ Ancestors of the Tapir from the Lower Mio- cene of Dakota.” 1 Osborn, American Naturalist, Sept., 1892, p. 763. 2 Bull. Am. Mus. Nat. Hist., Vol. IV, Sept., 1892, p. 127. 3 Bull. Am. Mus. Nat. Hist., Vol. V, August, 1893, pp. 159-180. 362 Bulletin American Museum of Natural History. (Vol. VU, Genus Colodon Marsh. There is no evidence that the true Hyrachyus-Hyracodon \ine existed in Europe; the Colodon genus or stage of Lophiodont development, is probably represented in France by the animal from St. Gérard de Puy, which Fihol has mistakenly referred to Hyrachyus,' as H. douvillet. In our former communication upon the American representa- » tives of this genus, we had no hesitancy in referring it to the family Helaletide from the North American Eocene, and regard- ing it as the probable successor of the Upper Eocene representa- tive (Helaletes) of this family. Additional material, collected by the Museum Expedition of last year, now enables us to not only clear up the question of the species, but at the same time throws a new light upon the probable family relationship of these ‘Tapir- oids, as above detailed. An analysis of the species may now be given as follows : Size large ; length of last two lower Ms. and last two lower Pms., .072. Postero-internal cusp of the last lower premolar double. Internal cusps of superior premolars not fully distinct ; no external nor internal cingula on premolars... ... Dye En Shes ee eee oe C. dakotensis. Size large ; length of lower Pms. and Ms. ahknows 4 last inferior pre- molar unknown. Internal cusps of second and third upper pre- molars distinct and well separated; an external and internal cingulum upon premolars............. 2; oss » iG. PROCUSPIAAlUs. Size small; length of last two Ms. and last two lower Pms., .055. Postero-internal cusp of last lower premolar single. Superior premolars: unkDOWIM sone or) dhe ej a eee .. C. occidentalis. Previously established upon foot characteristics only, possibly equivalent HOO WIAA USE IGG SO ORLA HaOr ADUNNo WoeaDONdooo aa occ C. longipes. Colodon dakotensis, sp. nov. The type of this species consists of an entire superior molar and premolar dentition lacking only the first premolar of the left side (No. 1212). ‘To this we add as a collateral type a specimen of another individual displaying the second and third lower pre- molars, the second and third lower molars of the right side, and the fourth upper premolar of the left side (No. 1213). 1 Annales des Sciences Geologiques, ‘T. xvii, pl. vi, fig. 13. 2 Wortman and Earle, ‘ Ancestors of the Tapir from the Lower Miocene of Dakota,’ Bull. Am. Mus. Nat. Hist., Vol. V, 1893, Art. XI, pp. 159-180. . i4 1895.| Osborn and Wortman, Perissodactyls of White River. 363 The superior cheek teeth consist of four premolars and three molars. The first premolar is small, having a triangular crown with a single fully-developed external and internal cusp. ‘The postero-external cusp (tritocone) is faintly indicated by a groove in the main external cusp, as is also the antero-internal (deutero- cone) represented by a small but distinct tubercle situated just in advance of the large internal cusp. Q paracone parastyle I f 3 ‘ metacone 7 N er Ns Ss ¢ > / x Fri metaloph feta rtocone deuterocone Fig. 7. Upper molar and premolar series of Colodon dakotensis, internal and crown views. Slightly larger than natural size. The succeeding three premolars increase slightly in size from before backwards ; their crowns are more or less quadrate in out- line, and each displays a double external and internal cusp con- nected by well-defined cross-crests. The internal cusps of the premolars are not fully developed and distinct from each other in this species, but are indicated by a deep vertical groove upon the internal face of the crown. It is a matter of importance to note that in the assumption of the double internal cusps of the pre- molars, this species furnishes us with the incipient and transition stages, and further, that this complication began in the second premolar and proceeded backwards. ‘This is demonstrated by the fact that the second premolar is more advanced in this respect than the third, and the third is more advanced than the fourth. The arrangement of the external cusps is somewhat different from that of the true molars, in that the posterior external cusps of the premolars are not pressed inwards and concave as they 364 Bulletin American Museum of Natural History. |Vol. V1I, are in the molars. ‘The parastyle at the antero-external angle of the crown is faintly but clearly indicated, and there can be said to be no external or internal cingula developed upon any of the premolars. The structure of the true molars has already been described,’ and, so faras can be determined from the materials at hand, varies but very little in the different species. It is, however, worthy of remark that the cingulum in this species is but faintly if at all indicated upon any of the molars. Of the inferior molar dentition the structure is very similar to that of C. occidentale in general appearance. An important struc- tural difference between the species, however, is to be seen in the last inferior premolar ; in C. dakotensis the posterior portion of the crown widens rapidly, and the postero-internal cusp is double, whereas in C,. occidentale this portion of the tooth is relatively much narrower and the cusp is single. Associated with difference of structure is a marked difference in size between the species ; C. dakotensis is larger and more robust in every way. This is made more apparent by a comparison of the following meas- urements: C. dakotensis. C. occidentalis. M. M. Length of last two lower molars .. .......... pare Onn .034 Length of last lower molar. ........... Sete Bes .025 -O19 Length of last two lower premolars......... .-... .027 .021 Width of crown of last lower premolar ........... .O10 .013 Rotallengthiotsupper molariseriess =. eevee oe .OgI = engsthvofipremolars/abover yer icles etre Od — This species is from the Metamynodon layer, and was found by Mr. O. A. Peterson, a member of the party. Colodon procuspidatus, sp. nov. This species is proposed upon a complete superior maxillary dentition of the right side, in which the last molar is wanting (No. 1215). So far as the measurements are concerned, it agrees very closely in size with C. dakotensis. The most important difference between this species and C. dakotensis is seen in the PLGC: Cit:.) Pet 75. 1895.] Osborn and Wortman, Perissodactyls of White River. 365 degree of separation of the internal cusps of the premolars from each other, and the more decided approach towards the structure of the molars. In the second premolar the two internal cusps are almost as distinct as they are in the molars; in the third premolar they are less so, while in the fourth they are separated by scarcely more than a vertical groove on the internal face of the paracone ‘parastyle / , , paras ty i a ¢ 3 metacone Fig. 8. Upper molar and premolar series of Colodon procuspidatus, internal and crown views. Slightly larger than natural size. crown. The cross-crests are more prominent than in the pre- molars of C. dakotensis, but the two external cusps are apparently not so distinct from each other as in that species. The external and internal cingula are prominent and distinct. The only means at present known of distinguishing C. procuspidatus from C. occidentale is by the smaller size and generally less robust char- acter of the latter. Found in the Metamynodon layer by Mr. J. W. Gidley. Lower Milk Molars of C. ocCIDENTALIS.—The inferior milk molar dentition of this species is represented in our collections by two fragmentary lower jaws (Nos. 1044 and 1044a). With the exception of the first milk molar, which agrees very closely in size and structure with its corresponding premolar, the remaining two teeth of this series are of a more advanced pattern. They resemble the true molars in that the posterior cross-crest is com- plete and quite as well developed as the anterior. In the perma- 366 Bulletin American Museum of Natural History. [Vol. VII, nent premolars the posterior crest is never complete, the heel of the tooth preserving its primitive arrangement of a separate external and internal cusp. The total length of this series slightly exceeds that of the corre- sponding premolars. Colodon (?) longipes O. & IV. SYN. Mesohippus longipes O. & W. It seems proper in this connection to again call attention to the specimen which we have described under this name.’ It is prob- able that it is the foot of a species related to Colodon, although it differs in some important particulars from the fragmentary materials which we already know of Colodon occidentale. In some respects it resembles the Horses, but at the same time it presents such striking differences from any known members of this series as to absolutely prohibit its reference to any of the Equide. These differences may be summarized as follows: (1) The continuity of the ectal and sustentacular facets of the astragalus, as in the Rhinoceroses and Hyracodons generally ; (2) the great vertical depth of the ectocuneiform; and (3) the articulation of meta- carpal IV with the ectocuneiform, thus excluding the contact between the cuboid and metacarpal III, an extremely constant and highly diagnostic feature of all the Horses. Its nearest prototype is apparently found in the foot of Hep/o- don calcwculus of the Wahsatch. The two astragali are very similar in their details of structure, and the whole foot is strik- ingly similar in the two forms. Unfortunately the ectocuneiform ; is not preserved in our specimen of A. calcicu/us. A comparison of the foot of C. Jongipes with that of 77riplopus amarorum Cope, reveals the closest similarity in all details of structure. There can be very little doubt therefore that C. Jongipes is the direct successor of some species of /Ve/aletes or Triplopus ; and whether the foot in question is to be associated with any of the known species of Colodon is still an open question. We have therefore retained the specific name, and have provisionally referred it to the genus Colodon. 1 Osborn and Wortman, Bull. Am. Mus, Vol. VI, 1804, Art. VII, p. 214. 1895.| Osborn and Wortman, Perissodactyls of White River. 367 Family HYRACODONTID. We insert here a description of the skull of Myrachyus from the Bridger Eocene, which is important in its bearing upon the relation of the primitive Hyracodonts to the true Aceratheres or Rhinoceroses. Hyrachyus agrarius Ze/d. The skull of this important species has been known hitherto only from specimens showing the upper and lower teeth, the jaws and the posterior portion of the occiput in the Leidy (Philadel- phia Academy) and Cope collections. The American Museum collection from the Bridger includes many parts of the skeleton and a nearly perfect skull and jaws (No. 1645), as represented in Figs. 9, 10 and 11. It was figured upon a very small scale on Plate II of our earlier paper. Dentition.—All the teeth are preserved excepting the upper incisors. ~The formula is typical, 3,4, 4,3. The swcisors are compactly placed, and decrease in size from the median to the outer pair. The median lower incisors (;;) are decidedly chisel- shaped or spatulate and nearly procumbent; the outer incisors (7a) are the smallest of the series, as well as the most erect and pointed. The upper canine is slightly larger than the lower ; both canines are vertically placed, laterally compressed and some- what incisiform, rather than of the typical canine form ; in fact they resemble a much enlarged lateral incisor. This is an impor- tant character. Upper Premolars.—The premolars in both jaws are simpler than the molars, or pm. Sori aise Cee er serene a Te oRG) 53 igey IDO REINS Ni ceo Gute b SS. TOOL Sceide eRe me ohaie onsen 3 96 1.16 PENG Nori ere ARS AME ASSIS e Re, us momar ca 10l4 20 Sactals (estimated)... tancetyes kine reese certo 544 .135 Caudals TEAL cal’, (alas sikh iiiestatva tayo acne ceeae erat eared 2 5% 735 Ribs, PSUR a a sheetcrkee esl econ ean cet eee Pederson te To) si 2355 BEHURID ede nore patie system te take ere i rani peieome hen esters ele: .76 UME sarees ME PAP RIN A Ue RETR Sires Acosta 28 81 1895.| _ Osborn and Wortman, Perissodactyls of White River. 375 ° The animal in life was over nine feet long, about three feet broad through the chest, and nearly five feet high, for it is prob- able that the anterior dorsal spines were longer than here repre- sented. The general impression is of a very large skull with formidable canine tusks, small but prominent eye-sockets, and very broad, flat skull. The fore and hind limbs are quite power- ful, but the metapodials are rather slender, especially in the manus. ‘The most distinctive feature of course is the four com- pletely functional digits, which widely separate this animal from the true Rhinoceroses. The chest has a well-rounded barrel, and the lower border of the abdomen must have been quite low. The anterior ribs are flat, but from the R.7 backwards they become rounded and rather slender. The skeleton has already been described in some detail.’ 1 * Fossil Mammals of the Lower Miocene White River Beds, Collection of 1892,’ Bull. Am. Mus., Vol. VI, July; 1894, p. 2090. Tae vate SAD sat ae = aoe a ~ ee ee

a INDEX. Mouse, Leaf-eared Cliff, 229. Long-tailed Meadow, 266. Maximilian’s Pocket, 266. Miller’s White-footed, 227. Missouri Grasshopper, 268. Mountain Meadow, 219. Rocky Mountain Jumping, 266. Rowley’s White-footed, 227. Silky Cliff, 226. Sonoran Harvest, 235. Sonoran White-footed, 229. Texan White-footed, 269. White-bellied Meadow, 219. Multituberculata, 11-15. Mus accedula, 183. egyptius, 190. agrarius americanus, 189. americanus, 189. arenarius, 183. arvalis nigricans, 189 aspalax, 184. capensis, 183. carolinensis, 108, 116, 118. citellus, Igo. cricetus, 183. cricetus niger, 183. decumanus, 260, 270. furunculus, 183. humilis, 108, 116. humulis, 108, 116, 117, 118. lecontii, 108, 116. lemmus sibiricus, 190, -lenze, 190. lenensis, Igo. leucopus, 189. maritimus, 184. messorius, 18Q. mexicanus, IgO. microcephalus, 184. migratorius, 183. minutus, 189. minutus flavus, 189. moschatus, 189. musculus, 236, 260, 270. myospalax, 184. nigricans, 190. pheeus, 183. pilorides fulvus, 189. rutilus minor, Igo. songaricus, 183. suslica, Igo. talpina, 183. tazamaca, 135. tschelag, Igo. virginianus, Igo. (Myotalpa) talpina nigra, 190. Muscovite, 342. » Muskrat, 183. Pallid, 256, 266. Musquash, 183. Mustela afra, 188. americana, 188. guianensis, 188. javanica, 188. laniger, 188. lutra canadensis, 188. melina, 189. pennanti, 158. zibellina americana, 188. zibellina nigra, 188. (Lutra) canadensis, 188. (Lutra) chilensis, 188. (Lutra) guianensis, 188. (Lutra) paraguensis, 188. Myocastor, 181, 182. coypus, 181, 183. zibethicus, 183. Myopotamus, 181, 182. Myotalpa, 181, 183. aspalax, 184. capensis, 183, 184. maritima, 184. myospalax, 184. talpina, 183, 184. talpina nigra, 183. typhla, 184. Myoxus africanus, Igo. inauris, Igo. Myrmecophaga jubata sima, 187. pentadactyla, 187. striata, 187. NASALIS nasalis, 186. nasica, 186. nasuus, 1386. Neoplagiaulax, 5. americanus, 7. Neotoma californica, 223. campestris, 260, 269. cinerea occidentalis, 224. cinnamomea, 331. floridana, 223. grangeri, 262, 270. mexicana, 221. micropus, 224. rupicola, 262, 270. Noctilio leporinus, 186. Nyctibius jamaicensis, 320. Nyctinomus brasiliensis, 246. nevadensis, 245. OCHETODON, 109, 115. humilis, 109, 120, 125. longicauda, Log, 129. megaiotis, 109. mexicanus, 109, 123, 135, 137. 384 INDEX. Ochetodon montanus, 109. sumichrasti, 109. Onychodectes, 5, 40. ? rarus, 8, 42. tissoniensis, 8, 40. Onychomys leucogaster, 260, 268. leucogaster pallescens, 225. pallescens, 225. torridus, 224. Oreodon culbertsoni, 164. Oryzomys cherriei, 329. Osborn, Henry Fairfield, fossil mammals of the Uinta Basin, Expedition of 1894, 71-105. Osborn, Henry Fairfield, and Charles Earle, fossil mam- mals of the Puerco Beds, Collection of 1892, I-70. Osborn, Henry Fairfield, and J. L. Wortman, Perissodactyls of the Lower Miocene White River Beds, 343-375. Oxyacodon, 25. apiculatus, 9, 25. Oxycleenus cuspidatus, g. Ovis ammon europzea, 192. canadensis, 258. cervina, 258, 263. europea, 192. montana, 258. musimon, 192. PACHY NA, 5. Palosyopinz, 82. Panther, 253. Pantolambda, 2, 5, 43. bathmodon, 9, 43. cavirictus, 9. Pantolamdide, 43. Paradoxodon rutimeyeranus, 9. Paradoxurus gallica, 188. hermaphroditus, 188. typus, 188. Paramys uintensis, 75, 81. Pentacodon inversus, 9. Periptichidz, 49, 52. Periptichine, 52. Periptychus, 47, 53. brabensis, 9, 54. coarctatus, 9, 54. rhabdodon, 9, 53. Perodipus chapmani, 214. richardsoni, 260, 262, 265. Perognathus apache, 216. bimaculatus, 216. conditi, 219. fasciatus, 262, 266. flavus, 215. Perognathus obscurus, 216. paradoxus, 260, 262, 266. pricei, 216. Peromyscus attwateri, 330. auripectus, 226. eremicus, 226. leucopus, 189. leucopus arcticus, 262, 268. leucopus deserticolus, 231. leucopus nebrascensis, 202,265 leucopus rufinus, 197. leucopus sonoriensis, 229,231. leucopus texanus, 260, 269. megalotis, 229. rowleyi, 227. rowleyi pinalis, 197, 227. Petauroides volans, 189. Petaurus norfolcensis, Igo. sciurea, 190. Petchary, Black-banded, 323. Peterson, O. A., on the geology of the Uinta Basin, 72-74. Phaéthornis guyi, 325. ; Phenacodontide, 49, 64. Phenacodus, 47. Phlegothontius, 296, 3IT. carolina, 298, 311. cingulatus, 298, 311. quinquemacuiatus, 297, 311. rusticus, 300, 311. Phoca chilensis, 187. groenlandica, 187. groenlandica nigra, 187. hispida quadrata, 187. laniger, 187. maculata, 187. mutica, 187. nigra, 187. punctata, 187. testudo, 187. Pholus, 288, 295. achemon, 289, 295. linnei, 291, 295. pandorus, 288, 295. vitis, 290, 295. Piranga hemalea, 323. Pitangus sulphuratus, 321. Plagiaulacide, I1. Plagiaulacinze, IT. Plagiaulax, IT. Plesiarctomys sciuroides, 75. Pocket-mouse, Apache, 216. Arizona, 216. Brown, 216. Condit’s, 219. Polymastodon, 2, 5, 11. attenuatus, 7, 12, 13. fissidens, 7, 12, I4. foliatus, 7, II. INDEX. 385 er a es — | Eh, arg 9 A ea selenodus, 7, I2, I4. taoénsis, 7, II, 13, 14. Polymastodontine, II. Poor-me-one, 326. Porcupine, Yellow-haired, 265. Potorous tridactylus, 189. Prairie-dog. (See Dog.) Price, W. W. (See Allen, J. A.) Primates, 15, 76. Prionodon maximus, 187. Procyon lotor hernandezii, 250. Protochriacus, 22. attenuatus, 8, 22. priscus, 8, 22. simplex, 8, 23. Protogonia, 64. Protogonodon, 5, 12, 67, 70. lydekkerianus, 9. pentacus, 9, 67. Protoreodon, 175, 176. Polymastodon latimolis, 7, 11, 12. | Rabbit, Mearns’s Cotton, 220. ' : Merriam’s Kangaroo, 213. Mexican Wood, 221. Plains Wood, 269. Richardson’s Kangaroo, 265. ' Reithrodon, 109, 115. australis, 328. carolinensis, 109, 116. humilis, 116. lecontii, 116. longicauda, 109, 129. megalotis, 107, 109, 125. mexicanus, 109, 135. montanus, 107, 109, 123. sumichrasti, 109, 135. Reithrodontomys, 107, 109, I15. aztecus, 109, I10, 125. arizonensis, I13, II5, 134, 235. costaricensis, I13, 115, 139. dychei, 112, 114, 120, 236. dychei nebracensis, 112, 114, parvus, 75. 122, 260. Proviverridz, 39. fulvescens, 113, 115, 138, 235. Prude, 186. humilis, 116. Pseudochirus peregrinus, 189. lecontii) 110; "512; 114; 116; Psittacotherium, 5, 42. 14I. aspasciz, 8. longicauda, I10, I12, 115,129, megalodus, 8. 134, 142. multifragum, 8, 42. longicauda pallidus, 113, 115, Ptilodus, 5. TBE eA Ss: medizevus, 7. trovessartianus, 7. Puerco Beds, fossil mammals of, I-70 ; stratigraphy of, 1. Puerco Fauna, mainly Mesozoic, 3; synopsis and vertical dis- tribution of, 7-10 ; systematic megalotis, 107, 110, 112, I14, 125, 141, 234. megalotis deserti, 112, I14, 127i 142: merriami, IIO, I12, II4, 11g. mexicanus, I13, I15, 135. mexicanus aurantius, 113, 115, description of, I1-70. 137, 143. Putorius, sp. ?, 255. mexicanus fulvescens, IIo, longicaudus, 262, 273. 138, 235. : mexicanus intermedius, 113, 115, 136. RapBit, Allen’s Jack, 201. montanus, 107, TIO, 112, (14, Arizona Jack, 202. ok Eastern Black-eared Jack, 264. White-tailed Jack, 264. Raccoon, Black-footed, 250. RKamphocelus jacapa, 321. Rangifer tarandus caribou, IgI. tarandus grcenlandicus, Igr. Rat, Arizona Cotton, 220. Bad Lands, 270. Banner-tailed Kangaroo, 212. Black Hills Wood, 270. Brown, 270. Chapman’s Kangaroo, 214. Cherrie’s Cotton, 329. Desert Kangaroo, 212. Fulvous Wood, 331. pallidus, r10, 131. sumichrasti, I10. Rhinocerotidz, 371. Rougette, Lesser, 186. SAGOINUS, I81. Sapajou gris, 166. Sapajus, 181. Sarcothraustes, 28. antiquus, 8, 29. bathygnathus, 8, 30. crassicuspis, 8. coryphzus, 8, 29. Scalops aquaticus, 189. aquaticus argentatus, 260, 273. 386 INDEX. Sciurus aberti, 244. zestuans, 190. aestuans fasciatus, 190. albipes, 190. arizonensis, 245. arizonensis huachuca, 197, 245. badjing, 190. bancrofti, 190. capensis, 190. guianensis, 190. hudsonicus dakotensis, 262, 27 his hudsonicus grahamensis, 244. hudsonicus mogollonensis, 243. niger, 190. niger albirostro, 190. niger cinereus, 1gO. niger ludovicianus, 260, 270. petaurista, 190. plantani, Igo. scrotalis, 190. sp. as 259. Vs variegatus minor, Igo. Virginianus, 190. (Microsciurus) alfari, 333. (Petaurus) norfolcensis, 190. (Petaurus) petaurista, 190. (Petaurus) petaurista niger, 190. (Petaurus) virginianus, 190. Sciuropterus volans, Igo. Sheep, Mountain, 258, 263. Shrew, Forster’s, 273. Sigmodon hispidus arizonz, 220. minimus, 220. Simia annulata, 186. antiquensis, 186. cinerea, 185. comosa, 186. dentata, 185. ferox, 185. ferruginea, 186. flavescens, 186. fulva, 186. hircina, 186. lar argentatus, 185. lar minor, 185. leucopheea, 185. mona, 185. nasalis, 186. nasica, 186. pileata, 185. - satyrus, 185. satyrus jocko, 185. satyrus pongo, 185. sublutea, 185. suillus, 185. sylvicola, 185. Simia (Cercopithecus) cethiops tor- quatus, 185. (C.) aygula monea, 185. (C.) badius, 186. (C.) capistratus, 186. (C. ) exquima, 186. .) fulvus, 186. .) fuscus, 186. .) hamadryas ursinus, 186, >.) hircinus, 186. .) luteolus, 186. .) nasuus, 186. -.) nictitans barbartus, 186. S.) regalis, 186. .) ruber albofasciatus, 185. .) ruber nigrofasciatus, 185. .) silenus albibartus, 185. ‘.) silenus purpuratus, 185. .) silenus tie-tie, 185. .) Sinicus pileatus, 185. (C.) talapoin niger, 185. (C.) veter albibartus, 185. (C.) viridens, 186, (Papio) cinerea, 185. (P.) cristata, 185. (P.) livea, 185. (P.) sylvicola, 185. _ (P.) variegata, 185. (Sagoinus) jacchus moschatus, 186. (Sapajus) capucinus albulus, 186, (S.) trepidus fulvus, 186. (S.) variegatus, 186. Siskin, Pine, 197. Siphneus, 184. Sitomys americanus arizonz, 229, 231. americanus rufinus, 197. auripectus, 226. rowleyi, 227. rowleyi pinalis, 197, 227. Skunk, Arizona, 250. Black-tailed Striped, 274. Little Striped, 252. Texas ova. Smerinthus, 313, 318. astylus, 316, 318. exceecatus, 314, 318. geminatus, 313, 318. myops, 315, 318. Sorex albipes, 189. arcticus, 189. arcticus cinereus, 189. ceeruleus, 189. ceerulescens, 189. carinatus, 189. constrictus, 189. forsteri, 262, 273. ananananacanaaaa INDEX. Sorex liricaudatus, 189. mexicanus, 189. micrurus, 339. quadricaudatus, 189. sp. ? 255. tetragonurus, 189. unicolor, 189. vulgaris, 189. . (Crocidura) czruleus, 189. Spalax, 184. microphthalmus, 184. typhlus, 184. Spermophile, Black Hills, 337- Hoary, 239. Line-tailed, 237. Pale Striped, 271. Round-tailed, 238. Small Striped, 337. Spermophilus cryptospilotus, 197. macrospilotus, 197. tereticaudus, 197. tridecemlineatus, 338. tridecemlineatus olivaceus, 337, 339. tridecemlineatus pallidus, 260, 2625 27016330; tridecemlineatus parvus, 337, 339. Sphecodina, 283, 294. abbotii, 283, 294. Sphenoccelus uintensis, 75, 98. Sphingidze, 275. Sphinx, 300, 311. canadensis, 304, 311. chersis, 303, 311. drupiferarum, 300, 311. eremitus, 304, 3il. gordius, 302, 311. kalmiz, 301, 311. lucitiosa, 302, 311. plebius, 305, 311. Spilogale gracilis, 252. interrupta, 260, 274. Squirrel, Abert’s, 244. Alfaro’s, 333. Arizona, 245. Huachuca, 245. Say’s Ground, 241. Western Fox, 270. Sukotyrus, 181. indicus, 187. Sus tajassu minor, 192. tajassu patira, 192. Synallaxis carri, 323. Synetheres mexicanus, 189. Systemodon, 359. 387 TALIGRADA, 43. Talpa flava, 189. flavescens, 189. fusca, 189. Tama-macame, IgI. Tamandua, 187. Tamias cinereicollis, 243. dorsalis, 241. lateralis, 241. minimus, 262, 271. quadrivittatus borealis, 262, Parfit pricei, 333. wortmani, 335. Tanager, Rufous, 323. Tarsier, 186. Tatou a neuf bandes, 187. Tatusia longicaudatus, 187. Taxidea taxus, 261, 274. taxus berlandieri, 256. Telmatotherium, 82, 84. cornutum, 75, 83, 90. cultridens, 83, 95. diploconum, 75, 83, 85. hyognathum, 75, 83, 87. megarhinum, 75, 83, 84. vallidens, 83, 87. validum, 82, 83, 94. (Lurocephalus) cultridens, 83. (Palazosyops) hyognathum, 83. - (P.) megarhinum, 83. Tetraclenodon floverianus, 8. Thamnophilus albicrissus, 324. cirrhatus, 325. major, 324. ; major albicrissus, 324. trinitatis, 325. Theretra, 287, 295. tersa, 287, 295. Thomomys cervinus, 203. fulvus, 205. talpoides, 262, 265. Thrush, White-throated, 322. Tillodonta, 4o. Titanotheriidz, 82, 346. Titanotherium, 346. dolichoceras, 349. elatum, 349. proutii, 347, 348. robustum, 346. (Brontops) robustum, 347,348. (Titanops) elatum, 349. Tricentes, 23. bucculentus, 8, 23. crassicollidens, 8. ? subtrigonus, 8. Trichechus manatus siren, 187. Trichosurus vulpecula, 189. Trigonolestes, 5. 388 INDEX. Triisodon, 28. biculminatus, 8, 28. heilprinianus, 8. quivirensis, 8. Triisodontide, 28. Triplopus, 75. obliquidens, 75. Triptogon, 281, 294. lugubris, 281, 294. Torbenite, 342. Tourmaline, 342. Tucan of Fernandez, 189. ‘Tyrannus melancholichus satrapa, 3218 Uinta Basin, fossil mammals of, 71-105 ; geology of, 72-74; the three faunal levels of, 75 ; succession of species in, 74. Uintatherium, 75. Urocyon cinereo-argenteus scottii, ce virginianus scottil, 255. Ursus americanus, 255. horribilis ?, 255. indicus, 189. VESPERIMUS americanus sonorien- sis, 231. Vespertilio americanus, 186. cephalotes melinus, 186. ciliolabrum, 262, 273. evotis, 240. labialis, 186. Vespertilio lucifugus, 249. melanorhinus, 248. nitidus, 248. pictus rubellus, 186. vampyrus helvus, 186. Vesperugo hesperus, 247. Vireo chivi agilis, 321. Viverra gallica, 188. maculata, 188, 189. nems, 188. nigra, 188. prehensilis, 188. Vulpes macrotis, 255. macrourus ?, 274. WEASEL, 255. Long-tailed, 273. Wolf, Gray, 254, 274. Woodpecker, Yellow-headed, 324. Worthless, 322. Wortman, J. L., field notes on the Puerco Beds, 1 ; on the osteol- ogy of Agriocherus, 145-178. (See also Osborn, Henry Fair- field. ) XENOTINE, 341. Xerus capensis, 190. inauris, Igo. | ZApus hudsonius, 181. princeps, 262, 266. Zetodon gracilis, 9. 7AGCON S42: . i aes a4 ee ea 4, a ea Win eae =f i ’ " ep t i ra Py = in a \ pl QH American Museum of natural 1 history, New York A4 Bulletin Val Biological & Medical Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY acer tape wa ween ee anh Pe tame Oe 60 em en = Sade ae > anv Seana hah ate ihe tiepaan > . > 7 - » = ota : pate . . ase ered ie a me eg a wig ge ee a a nds eer Se eedeath th eee OE ey ~